思步网

标题: 【连载】CB Predictor操作手册(Crystal Ball Predictor 水晶球预测) [打印本页]

作者: lanj 时间: 2009-11-27 14:29
标题: 【连载】CB Predictor操作手册(Crystal Ball Predictor 水晶球预测)
从去年到上半年，一直在准备CMMI-4的评审。也因此接触了一些很不错的软件和统计工具，现在给大家介绍一种。Crystal Ball里面的CB Predictor预测工具。希望对大家都工作能有所帮助。

PS:由于是自己对英文教程的简单翻译，如有不足的地方，请大家见谅，凑合的看看啊～！

==============2010.07.19 管理员添加备注【开始】===============

该书的中英对照版电子书已发布，免费下载地址：http://step365.com/thread-2452-1-1.html

==============2010.07.19 管理员添加备注【结束】===============

作者: lanj 时间: 2009-11-27 14:31
CB Predictor™ 1.6—— User Manual使用手册
欢迎使用CB Predictor™

Welcome to CB Predictor, a powerful addition to the Crystal Ball suite of decision intelligence products!
欢迎使用CB Predictor，这是一个水晶球系统的附加的一个强大的决策产品！

Forecasting is an important part of many business decisions. Every organization needs to set goals, try to predict future events, and then act to fulfill the goals. As the timeliness of market actions becomes more important, the need for accurate planning and forecasting throughout an organization is essential to get ahead. The difference between good and bad forecasting can affect the success of an entire organization.
在商业决策中，预测是一个重要的部。每个企业都需要设立自己的目标，尝试去预测未来的发展趋势，然后采取行动去实现目标。随着及时的市场行动变的更为重要，这就需要我们在整个过程中能更准确的进行规划和预测。好的与坏的预测之间的差别是好的预测可以影响到一个组织的成功。

CB Predictor is an easy-to-use, graphically oriented forecasting add-in for Microsoft Excel spreadsheet users. If you have historical data in your spreadsheet, CB Predictor analyzes your data for trends and seasonal variations. It then predicts future values based on this information. You can answer questions such as, “What are the likely figures for next quarter’s sales?” or, “How much material do we need to have on hand?” With CB Predictor, you no longer need to pull these numbers out of thin air. Instead, you can rely on robust, statistically proven techniques to create these predictions accurately.
CB Predictor是一个易于使用的，自动加载在EXCEL中的预测向导图。如果在sheet中有历史数据，CB Predictor对数据的趋势和周期性变化进行分析，在历史数据的基础上预测未来的数值。您可以回答以下问题，如“下个季度的销售额可能是多少？”或者，“我们需要有库存在手头上？”使用 CB Predictor，你不需要再凭空的进行预测，你可以依靠完善的统计技术来进行准确的预测。

To get started, all you need is a spreadsheet containing historical data. From there, this manual guides you step by step, explaining forecasting terms and results.
若要开始使用CB Predictor，所有你需要一个包含历史数据的电子表格。本手册将引导您一步一步，解释条款和预测结果。

--------------------------------------------------------------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-11-27 14:31
Who should use this program?
谁会使用这一工具?

CB Predictor is for the planner and forecaster in every organization, from the sales manager predicting next quarter’s sales, to the marketing specialist forecasting the results of an advertising campaign, to the financial officer projecting likely revenue figures and cash flows. CB Predictor has been developed with a wide range of forecasting applications in mind.
CB Predictor对组织中计划和预测人员，对下一季度的销售量进行预测，对预测结果来预测广告投入量，和财务的现金收入和支出情况。

You don't need highly advanced statistical or computer knowledge to use CB Predictor to its full potential. All you need is a basic working knowledge of your computer and Microsoft Excel, plus a solid understanding of your own business fundamentals.
你不需要有非常先进的统计或计算机知识，利用CB Predictor充分发挥其潜力。您需要的是一些计算机和Microsoft Excel的基本知识，扎实的业务基础。

As an added benefit, you can automatically save CB Predictor forecasts as Crystal Ball assumptions for immediate use in powerful risk analysis models. See Chapter 5 for more information.
作为一个附加的好处是，可以直接对风险分析模型的建设自动生成CB Predictor预测。见第5章以获取更多信息。

CB Predictor runs on several versions of Microsoft Windows and Microsoft Excel. For a list of required hardware and software, see README.htm in the main Crystal Ball installation folder (by default, C:\Program Files\Decisioneering\Crystal Ball 7).
CB Predictor在多个版本的Microsoft Windows和Microsoft Excel中上运行。所需的硬件和软件，在README.htm中，主要水晶球安装文件夹（默认情况下， C：\ Program Files\ Decisioneering \ Crystal Ball 7)。

-------------------------------------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-11-27 14:34
How this manual is organized
手册是有哪些内容组成的？

•Chapter 1 — “Getting Started with CB Predictor”
启动 CB Predictor

This chapter contains two tutorials designed to give you a quick overview of CB Predictor’s features and to show you how to use it. Read this chapter if you need a basic understanding of CB Predictor.
本章包括两个案例来帮助你对CB Predictor的特性和如何使用进行快速的了解。

•Chapter 2 — “Understanding the Terminology”
了解术语

This chapter contains a thorough description of forecasting and statistical terms. Read this chapter carefully if your statistics background is limited or if you need a review of how terms are used in CB Predictor.
本章描述了预测和统计的条件。

•Chapter 3 — “Forecasting with CB Predictor”
使用 CB Predictor进行预测

This chapter contains step-by-step procedures for using all the features in CB Predictor.
本章描述了CB Predictor预测的一步一步的步骤和所有的特性。

•Chapter 4 — “Examples”
案例
This chapter contains example data from various fields.
本章包含各个领域的数据案例。

•Chapter 5 — “Using CB Predictor with Crystal Ball”
使用Crystal Ball 中的CB Predictor
This chapter contains descriptions of how to use CB Predictor with Crystal Ball. It also has some examples that are Crystal Ball-specific. The manual includes the following:
本章说明如何使用CB Predictor与水晶球。它还包括一些案例：

•Appendices
附录

•A — “The CB Predictor Wizard”
Settings for each tab of the CB Predictor wizard dialog.
A——CB Predictor 向导
CB Predictor向导每项步骤的设置

•B — “Time-Series Forecasting Method Formulas”
The formulas used for the time-series forecasting methods.
B——时间序列预测方法公式
使用时间序列预测方法的公式有哪些

•C — “Error Measure and Statistic Formulas”
The formulas used for time-series error measures and other statistics.
C——误差度量和统计的公式
使用时间序列预测方法的公式对误差和其他方面进行统计度量

•D — “Techniques for Finding Regression Coefficients”
How the program determines the coefficients of regression equations
D——回归系数的统计方法
如何确定回归方程中的回归系数

•E — “Regression Statistic Formulas”
The formulas used to calculate the statistics to gauge the quality of the regression and the coefficients.
E——回归统计公式
回归方程和回归系数的计算公式和统计方法

•F — “Error Messages”
A list of the most important CB Predictor error messages with directions for error recovery.
F——误差信息
罗列了CB Predictor中的几种重要的误差种类

•Bibliography
A list of related publications, including statistics textbooks.
•参考书目
有关的出版物的清单，包括统计教科书

•Glossary
A compilation of terms specific to CB Predictor as well as statistical terms used in this manual.
术语
CB Predictor方面的相关术语，以及在本手册中使用的统计术语汇编。

•Index
An alphabetical list of subjects and corresponding page numbers.
序号
手册和相应的页码内容字母排列顺序

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: 漂在生活 时间: 2009-11-27 15:14
楼主好强啊，顶！！！
期待楼主在连载的同时，也能教教想学的同学们。。。

作者: step365he 时间: 2009-11-30 09:25
楼主翻译的效率很高，质量又好~

作者: lanj 时间: 2009-11-30 09:26
{:8_305:}听着怎么感觉毛毛的～
看来要加油了，否则对不起大家啊{:8_307:}

作者: lanj 时间: 2009-11-30 09:29
Chapter 1 第一章
Getting Started with CB Predictor

In this chapter在这一章中会涉及到：
•What is time-series forecasting?
   什么是时间序列预测？
•Shampoo Sales tutorial
   香波销售案例
•How CB Predictor works
   CB Predictor如何工作
•Toledo Gas tutorial
   托莱多燃气公司案例

This chapter has two tutorials, a short one and a longer one, that provide an overview of CB Predictor’s features. The first tutorial, Shampoo Sales, is ready to run, using most of the defaults. The results forecast the next quarter’s worth of shampoo sales。
本章有两个案例，一个简单的一个复杂的，这两个案例是帮助我们更好的对CB Predictor的特性进行了解。在第一个香波案例中，有许多的误差，它是对下一个季度香波的销售值的预测。

The second tutorial, Toledo Gas, uses regression, previews the results, and customizes some features. The results predict the next year’s worth of residential gas usage for Toledo.
第二个Toledo Gas的案例，使用了回归分析，对结果进行评审，是对Toledo Gas明年的使用情况的预测.

Now spend some time learning how CB Predictor can help you forecast the future。
现在花费一些时间来学习如果使用CB Predictor来帮助你预测未来。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-11-30 09:30
What is time-series forecasting? 什么是时间序列预测？
“Prediction is very difficult, especially if it’s about the future.”- Niels Bohr Nobel laureate in Physics
“预测是非常困难的，尤其是对未来的预测”——Niels Bohr 诺贝尔物理学奖获得者

How often are you asked to estimate future sales? Staffing requirements? Budget items? Do you need to predict numbers for the next year using historical numbers that are higher in the summer than the winter? Or numbers that rise and fall with the Dow Jones industrial average? if so, CB Predictor can help.
多久询问一次未来的销售量？员工配置需求？财政预算？你是否需要用比冬季更高的夏季的历史数据来预测明年的数据？或者是道琼斯工业平均指数的上涨和下跌？如果是的话，CB Predictor可以帮助你。

If you have historical time-series data, you can use CB Predictor to examine your historical data and predict future trends. You no longer have to assume that sales will be same as last quarter or that you will spend 10% more next year on expenses than this year. CB Predictor applies a battery of sophisticated statistical methods to your data series to find results that meet the strictest confidence requirements.
如果你有有时间顺序的历史数据，你可以使用CB Predictor工具来对历史数据进行评估并对未来的趋势进行预测。你不需要再去假设你的销售额会和上个季度的一样，也不需要再花费比去年多10%的开支。CB Predictor工具有一整套成熟的统计方法对你的历史数据进行分析，并根据严格的要求得到预测的结果。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-11-30 09:31
CB Predictor uses two types of forecasting: time-series forecasting and multiple linear regression. Time-series forecasting breaks down your historical data into four components: level, trend, seasonality, and error. CB Predictor analyzes each of these components and then projects them into the future to predict likely results.
CB Predictor使用两种类型的预测方法：时间序列预测和多元线性回归方法。时间序列预测方法将你的历史数据从四个方面进行分析：水平、趋势、周期性和误差。CB Predictor对每个方面进行分析，并利用他们来预测未来最有可能的结果。

When you know that outside influences have an effect on the variable you want to forecast, you need regression. Regression takes historical data from the influencing variables and determines the mathematical relationship between these variables and your target variable. It then uses time-series forecasting methods to forecast the influencing variables and combines the results mathematically to forecast your target variable.
当你知道需要预测的变量收到外部因素的影响时，在预测是需要进行回归。回归是要对找到有影响因素的历史数据，并找到这些因素和预测变量之间的相互数学关系，然后运用时间时间序列预测的方法结合变量和目标变量之间的相互关系来进行预测。

After finding the best forecast for your data, CB Predictor pastes the forecasted values into your spreadsheet. You can also request detailed output that includes statistics, charts, reports, and PivotTables. CB Predictor can also create your forecasted values as Crystal Ball assumptions, ready for a “what-if” simulation.
当从数据中找到了最佳的预测值收，CB Predictor将预测结果粘贴到sheet表格中。你也可以要求详细的输出结果，包括统计数据、图表、报告和Pivto图。CB Predictor预测的数据可以作为水晶球的假设，为准备模拟做准备。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: 思步 时间: 2009-11-30 09:35
每天等待更新，就跟等待吃饭一样，是一种享受啊，急切！

作者: lanj 时间: 2009-12-1 09:13
本帖最后由 lanj 于 2009-12-1 09:31 编辑

Shampoo Sales tutorial香波销售案例
The easiest way to understand what CB Predictor does is to apply it to a simple example. In this example, you are sales manager for Tropical Cosmetics Co. The company’s latest product, shampoo with tropical ingredients, has been in the marketplace for almost a year now. The vice president of marketing wants you to forecast the rest of the year’s sales of shampoo and decide whether to recommend investing in advertising or enhancements for this product.
CB Predictor最简单的理解方法就是进行案例的演示。在这个案例中，你是Tropical Cosmetics Co.公司的销售经理，这家企业最新的产品是有热带成分的洗发香波，该产品已经在市场是销售了近一年的时间。负责营销的副总希望你对今年剩下的时间的销售量进行预测，以便来决定是否在广告投入和产品改良中继续投资。

You have the weekly sales numbers for the last nine months.
你有过去9个月的每周的销售数据。

To begin the tutorial: 案例开始：
1.Start Crystal Ball and Excel.
1、打开水晶球（crystalball）和excel表格。
2.Open the Shampoo Sales spreadsheet from the Examples folder.
2、在案例文件夹中打开香波销售量的电子表格。
By default, the file is stored in this folder: C:\Program Files\Decisioneering\Crystal Ball 7\Examples\CB Predictor Examples.
该表格的默认地址是C:\Program Files\Decisioneering\Crystal Ball 7\Examples\CB Predictor Examples.

ps:我将Shampoo sales的表格上传给大家看看

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

shampoo sales.png (127.45 KB, 下载次数: 190)

Shampoo Sales.xls

355 KB, 下载次数: 11349

作者: lanj 时间: 2009-12-1 09:18
In this spreadsheet, there is one column of Tropical shampoo sales data next to a column with dates from January 1, 2004 until September 23, 2004. You need to forecast sales through the end of the year, December 31, 2004.
在这张电子表格中，一列是热带香波的销售量的数据，边上一列是数据对应的时间，从2004年1月1日到2004年9月23日。你需要预测的是从现在到年底2004年12月31日的销售数据。

3.Select cell B4.
Selecting any one cell in your data range, headers, or date range initiates CB Predictor’s “Intelligent Input” to select all the filled, adjacent cells.
3、选择单元格B4
选择表格范围中任意一个单元格，以选择整个数据表格。

4.Select Run > CB Predictor.
4、选择Run>CB Predictor
The CB Predictor dialog, or wizard, opens to the Input Data tab as shown in Figure 1.2.
出现CB Predictor的对话框，按照图1.2显示的选取输入的数据范围。

Note: This command is only available if no simulation is running and the last run was reset. If necessary, wait for a simulation to stop or reset the last simulation.
注：此命令仅适合没有进行模拟运行的数据，如果有，需要将模拟停止或重新进行重置。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

输入数据选项卡.png (202.44 KB, 下载次数: 172)

Figure 1.2 CB Predictor wizard, Input Data tab

作者: 漂在生活 时间: 2009-12-1 12:51
又更新了，继续学习ing。:dabin

作者: zye0509 时间: 2009-12-1 16:10
定一下~~~~~~~~~~~~~~~~

作者: qicai2002 时间: 2009-12-1 17:49
谢谢分享，很需要啊！

作者: lanj 时间: 2009-12-2 09:46
When you select any one cell in the data range before you start the wizard, CB Predictor’s Intelligent Input guesses:
当你选择了数据表格中的任意一个单元格后再开始向导，CB Predictor会自动选择输入的范围：

•Your data series (in this case, A3:B42)
你的数据系列是在A3:B42的范围内

•Whether your data are in columns or rows
数据是按列或者按行排列

•Whether you have headers at the beginning of your data
你是否需要在开头开始数据

•Whether your first column or row contains dates or time periods
在第一列或第一行中是否包括数据活时间周期

5.Click Next.
5、点击Next

The Data Attributes tab appears as shown in Figure 1.3.
数据会像图1.3一样显示在向导卡中

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.3.png (93.91 KB, 下载次数: 68)

Figure 1.3 CB Predictor wizard, Data Attributes tab

作者: lanj 时间: 2009-12-2 09:53
6.Under Step 4:
a. Select “weeks” from the Data Is In list.
b. Set the data to have no seasonality.
You have less than two complete seasons (cycles) of data, so cannot use seasonality.
6、第四步：
a、在Data is in表格中选择“周”
b、选择没有周期性
如果选择周期性的话，你必须要有两个以上的完整周期的数据。

7.Under Step 5, make sure that Use Multiple Linear Regression is not checked.
You did not choose regression because you have only one series of data, so there are no dependencies between series requiring regression.
7、第五步，确认没有选择多元线性回归分析
如果只有一组数据的话，不需要选择线性回归，

8.Click Next.
8、点Next
The Method Gallery tab appears as shown in Figure 1.4.
图1.4显示了有哪些统计方法：

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.4.png (104.23 KB, 下载次数: 64)

Figure 1.4 CB Predictor wizard, Method Gallery tab

作者: lanj 时间: 2009-12-2 09:56
9.Click Select All.
This selects all the time-series forecasting methods, but CB Predictor doesn’t use the seasonal methods, since you indicated that your data were not seasonal.
CB Predictor forecasts your values using each of the selected methods and ranks them according to how well they fit the historical data. CB Predictor uses the seasonal methods as well as the nonseasonal methods if you indicate on the Data Attributes tab that your data series have seasonality.
9、点击“select all”
选择了所有的时间序列的方法，但是如果你的数据没有周期性，CB Predictor不使用周期性方法。
CB Predictor使用每个选择的方法来预测数值，并将产生预测值的最佳方法进行排序，如何你的数据系在数据分布图中显示的是有周期性的，那么CB Predictor也会选择周期性的方法，而不是非周期性的。

10.Click Next.
The Results tab appears as shown in Figure 1.5. The only output selected by default is Paste Forecast, which adds the forecasted values to the end of your historical datas shown in Figure 1.5.
10、点Next
在图1.5中显示了结果的图表，唯一输出的结果是粘贴预测，这种预测方法会将预测值粘贴到历史数据的最后面。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.5.png (86.95 KB, 下载次数: 61)

Figure 1.5 CB Predictor wizard, Results tab

作者: 漂在生活 时间: 2009-12-2 11:23
每天等更新……楼主加油哦！

作者: lanj 时间: 2009-12-3 09:28
11.Under Step 7, forecast the weekly sales for the rest of the year by entering 13 in the field.
11、第七步，在表格中填入需要预测的今年剩下的时间的销售量的周数，输入13周。

12.Click Preview.
The Preview Forecast dialog appears. It presents a graph with historical data, fitted data, forecast values, and confidence intervals as shown in Figure 1.6.
12、点击Preview
Preview Forecast对话框出现，这张图表中显示了历史数据、合适的数据、预测数据和置信区间的范围。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.6.png (189.49 KB, 下载次数: 75)

Figure 1.6 Preview Forecast dialog

作者: lanj 时间: 2009-12-3 09:33
13.In the Preview Forecast dialog, click the Method field.
The field lists all the methods CB Predictor tried, in order from the best-fitting method (designated by the word “Best”) to the worst-fitting method. CB Predictor calculates the forecasted values from the method that best fits the historical data. In this case, the method is Double Exponential Smoothing.
13、在Preview Forecast对话框中点击Method field。
在fiels清单中有CB Predcitor所有的预测方法，按照最佳到最差的方法顺序进行排列。CB Predictor从最佳的方法中历史数据情况计算出预测值，该图中最佳的方法是双边指数平滑。

The forecasted values appear as a blue line extending to the right of the historical data (green) and the fitted values also in blue). Above and below the forecasted values is the confidence interval (in red), showing the 5th and 95th percentiles of the forecasted values.
绿色的线代表的是历史数据，历史数据的最佳值对应的是与历史数据重合的蓝色的线条，在边上的蓝色线条代表的是预测值的趋势情况。在预测值的蓝色线条上下的红色线条是置信区间的范围，即预测值的5%-95%的可能性。

14.Click Run.
The program pastes the forecasted values at the end of the historical data (in bold), extending the date series as well. The forecasted values were forecasted using the best method, as shown in the Preview dialog.
Figure 1.7 Pasted shampoo sales values
14、点击Run
在历史数据后面粘贴预测值，该预测值是在Preview Forecast图中显示的最佳的方法得到的。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.7.png (114.13 KB, 下载次数: 58)

Figure 1.7 Pasted shampoo sales values

作者: lanj 时间: 2009-12-3 09:36
Based on the results, you complete your memo to upper management. Current strategies seem to be working so you recommend funding another project instead.
在这个结果的基础上，你根据上级管理层的生成备忘录，你应该建议改变策略，投资其他的产品。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.8.png (66.22 KB, 下载次数: 48)

Figure 1.8 Analysis memo

作者: lanj 时间: 2009-12-3 09:38
以上是用CB Predictor进行一个简单的销售量的预测，明天会将CB Predictor如何进行工作的相关内容与大家分享

作者: lanj 时间: 2009-12-4 09:02
How CB Predictor works CB Predictor如何工作？

Most historical or time-based data contain some kind of underlying trend or seasonal pattern. However, most historical data also contain random fluctuations — “noise”— that make it difficult to detect these trends and patterns without using a computer. CB Predictor uses sophisticated time-series methods to analyze the underlying structure of your data. It then projects the trends and patterns to predict future values.
大部分的历史或时间基础的数据都包含了相关的趋势和周期性，但是有些历史数据也是随机的，如果不借助电脑很难找到它的趋势性，CB Predictor使用时间序列趋势的图表方法对数据的结构进行分析，从而得到预测值的趋势和模型。

When you run CB Predictor, it tries each time-series method on your data and calculates a mathematical measure of goodness-of-fit. CB Predictor selects the method with the best goodness-of-fit as the method that will yield the most accurate forecast. CB Predictor performs this selection automatically for you, but you can also select individual methods manually or override the method CB Predictor recommends with a different one.
当你运行CB Predictor，它对历史数据运用各种时间序列的方法进行计算，得到一个最佳的数学度量。CB Predictor选择最佳的方法从而得到最精确的预测值，CB Predictor将为你自动选择最佳的方法，当然你可以根据自己的要求在方法图表中选择其他的。

The final forecast shows the most likely continuation of your data. Keep in mind that all these methods assume that some aspects of the historical trend or pattern will continue into the future. However, the farther out you forecast, the higher the likelihood that events will diverge from past behavior, and the less confident you can be of the results. To help you gauge the reliability of your forecast, CB Predictor provides a confidence interval indicating the degree of uncertainty around your forecast.
在你的数据上显示了最有可能的预测值，要记住，这些方法都是假定历史数据的趋势和结构将会延续到将来。但是，预测的值的时间越长，就会越有可能偏离以前的行为，预测的可信性也就越低。为了保证预测值的可靠性，CB Predictor提供了一个置信区间，通过这个置信区间来表示预测值的不可信度。

The final step of forecasting involves interpreting the results and integrating them into your decision-making process. CB Predictor provides comprehensive output to assist you in this process. You can request detailed reports, customizable Excel charts and graphs, and summary PivotTables. CB Predictor can also automatically create Crystal Ball assumptions for each forecasted value, letting you integrate risk analysis into the process.
在预测最后一步是解释结果，并将结果整合到您的决策过程中。文件提供全面的销售量预估，可以帮助您完成这一过程。您可以要求得出详细的报告，可定制的Excel图表和图表，数据透视表和汇总。CB Predictor还可以自动为每个水晶球预测值进行假设，帮你将入风险分析也考虑进来。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-4 09:03
本帖最后由 lanj 于 2009-12-4 09:09 编辑

Toledo Gas tutorial托莱多燃气公司案例

Suppose you work for Toledo Gas Company in the Residential Division. The Public Utilities Commission requires you to predict the gas usage for the coming year to make sure that the company can meet the demand.
假设你在托莱多住宅天然气公司工作，现在公共事业委员会要求你对明年的燃气使用量进行预测，以确保公司能够满足市民的需求。

To start the tutorial: 案例开始：

1.In Excel with Crystal Ball loaded, open the Toledo Gas spreadsheet, Toledo Gas.xls.
To find this example, select Help > Crystal Ball > Crystal Ball Examples and locate it in the list of CB Predictor examples, or browse to its folder. By default, the file is stored in this folder:
C:\Program Files\Decisioneering\Crystal Ball 7\Examples\CB Predictor Examples.
1、打开托莱多天然气电子表格，默认的存放位置是：C:\ProgramFiles\Decisioneering\Crystal Ball 7\Examples\CB Predictor Examples.

When you double-click the link or the file, the Toledo Gas spreadsheet appears in Excel.
双击文件夹后，托莱多燃气公司的数据表会打开

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.9.png (167.57 KB, 下载次数: 55)

Figure 1.9 Toledo Gas spreadsheet

作者: lanj 时间: 2009-12-7 09:30
2.Select cell C5.
2、选择单元格C5

3.Choose Run > CB Predictor.
The Input Data tab appears. CB Predictor’s Intelligent Input feature selects all the data from cell B4 to cell F100.
3、选择Run>CB Predictor
输入数据的表格出现，CB Predictor自动输入所有的数据，从B4：F100。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.10.png (193.5 KB, 下载次数: 57)

Figure 1.10 Input Data tab for Toledo Gas model

作者: lanj 时间: 2009-12-7 09:31
本帖最后由 lanj 于 2009-12-7 09:33 编辑

4.Click Next.
The Data Attributes tab appears
4、点Next
出现数据分布对话框

5.Confirm that in Step 4, settings are “Data is in months with seasonality of 12 months.”

5、检查第四步，设置数据输入的周期性是12个月。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.11.png (88.9 KB, 下载次数: 50)

Figure 1.11 Data Attributes tab

作者: lanj 时间: 2009-12-7 09:34
Using regression使用回归
Through research, you know that your residential gas usage is primarily affected by three variables: new home starts, the temperature, and the price of natural gas. However, you aren’t sure how much effect each has on gas usage.
通过调查，可以知道住在天然气使用量是收到三个方面因素的影响：新房的数量、问题和天然气的价格。但是你不能确认天然气是使用效果是多大？

Because you have independent variables affecting a dependent variable (the variable that you are interested in), this forecast requires regression. For regression, CB Predictor uses a technique called HyperCasting™, which in one easy step:
因为对因变量影响的有多个自变量，预测的结果需要进行回归。用回归的方法，CB Predictor使用一种叫做HyperCasting的技术：

a. Creates an equation that defines the mathematical relationship between the independent variables and your dependent variable.
a、根据因变量和自变量之间的关系建立一个数学方程式。

b. Forecasts each independent variable using time-series forecasting methods.
b、使用时间序列预测方法预测每一个因变量

c. Uses the equation it created in the first step, combining the forecasted independent variable values, to create the forecast for the dependent variable.
c、使用第一步创建的方程式，结合自变量的预测值，生成因变量的预测值。

In the Toledo Gas spreadsheet, the dependent variable is the historical residential gas usage. The independent variables are:
•Number of occupancy permits issued (new housing completions)
•Average temperature per month
•Unit cost of natural gas
在托莱多天然气电子表格中，因变量是居民天然气使用量的数据，自变量是：
       允许使用的数量（新房的居住率）
       每个月的平均气温
       天然气的单位成本
To resume your tutorial:

6.Under Step 5, select Use Multiple Linear Regression.
6、在第五步之后，选择多元线性回归。
The Regression Variables dialog appears.
回归变量的对话框显示出来

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.12.png (132.3 KB, 下载次数: 58)

Figure 1.12 Regression Variables dialog

作者: lanj 时间: 2009-12-7 09:36
7.Confirm that Usage appears in the Dependent Variables list.
If it does not:
•Select Usage in the All Series list.
•Click >> next to the Dependent Variables list.
The Usage variable moves to the Dependent Variables list.
7、确认在因变量列表中有“使用量”这个因素
如果没有：在所有系列列表中选择“使用量”，点击chick移到因变量列表中。

8.Confirm that Occupancy Permits, Average Temperature, and Cost Of Nature Gas Per Ccf appear in the Independent Variables list.
If they do not:
•Select Occupancy Permits, Average Temperature, and Cost Of Natural Gas Per Ccf in the All Series list.
•Click >> next to the Independent Variables list.
The variables move to the Independent Variables list.
8、确认允许使用的数量（新房的居住率）、每个月的平均气温、天然气的单位成本是否在自变量列表中。
如果没有，选择自变量，移到自变量列表中。

9.Click OK.
The Data Attributes tab reappears.
9、点击OK。出现数据分布图对话框

10.Click Next.
The Method Gallery tab appears.
10、点解Next。出现概率分布图对话框。

11.Click Next again.
The Results tab appears.
11、点击Next。出现结果。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: 漂在生活 时间: 2009-12-7 10:13
mark too.

作者: lanj 时间: 2009-12-8 09:33
本帖最后由 lanj 于 2009-12-8 09:34 编辑

Selecting results选择结果

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.13.png (76.47 KB, 下载次数: 51)

Figure 1.13 Results tab, set as described below

作者: lanj 时间: 2009-12-8 09:37
本帖最后由 lanj 于 2009-12-8 09:39 编辑

CB Predictor has five ways you can output your results:
CB Predictor 有五种结果输出的方法：

Paste Forecast

粘贴预测

Pastes the forecasted values to the end of your historical data.

将预测值粘贴到历史数据的最后面。

Charts

图表

Graphs historical, fitted, and forecasted data values, and a confidence interval.

将历史数据、历史最佳值、预测值和置信区间用图表表示出来。

Report

报告

Organizes and displays summary information, forecast values and a confidence interval, charts, and method information for any or all of your data series.

将数据的预测值、概要信息、置信区间、图表和方法形成报告。

Results Table

结果列表

Creates a table with all the forecasted values, fitted data, and a confidence interval.

将历史数据、历史最佳值、预测值和置信区间用结果列表表示出来。

Method Table

方法列表

Creates a table listing all the methods tried, the error values and statistics for each, and the parameters for each.

将所有的方法的误差值、统计值和参数用表格列出来。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-8 09:44
本帖最后由 lanj 于 2009-12-8 09:47 编辑

Paste Forecast and Charts are the most common results. The other settings provide more detailed output that includes information you might use for presentations or to investigate various forecasting methods.
粘贴预测和图表预测是最通常的表示结果，其他的方法提供详细的输出结果，包括你可以使用的用来预测和报告的各种方法。

This tutorial produces more detailed output. To continue with the tutorial:
这个案例会得出许多详细的结果。
12.Under Step 7, forecast the monthly usage for the next year by entering 12 in the field.
12、在第七步的基础上，在表格中输入需要预测的明年12个月的天然气使用量。

13.Select the Paste, Report, and Methods Table result settings.
13、选择粘贴、报告和方法列表的输出结果。

14.In the Title field, type “Toledo Gas Usage Forecast.”
Now, the dialog looks like Figure 1.13 on page 21
.14、输入标题。

15.Click Preferences
15.点击参数

The Preferences dialog appears as shown in Figure 1.14.
图1.14显示了参数设定对话框

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.14.png (58.38 KB, 下载次数: 60)

Figure 1.14 Preferences dialo

作者: lanj 时间: 2009-12-8 09:50
Notice that Paste Forecasts... is checked. This prepares forecasted independent variable data for use as assumptions in Crystal Ball risk analysis simulations.
注意粘贴预测中，因变量预测的数据是为了之后用水晶球进行分析，形成模型。

16.Select the Report tab.
16、选择报告
The Report preferences appear.
结果参数表会显示出来：

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.15.png (69.42 KB, 下载次数: 53)

Figure 1.15 Report Preferences dialog

作者: lanj 时间: 2009-12-8 09:54
17.Under Methods, select Three Best from the list and unselect the Parms (Parameters) setting.
This removes the method parameters from the report.
17、在方法列表中，选择三个最好的列表，不要选择设置参数。

18.Click OK.
The Results tab reappears.
18、点击ok。结果显示出来

19.Click Preview.
The Preview Forecast dialog appears.
19、点击Preview。预测结果预览对话框显示出来。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: 漂在生活 时间: 2009-12-8 11:30
再一次mark...

作者: lanj 时间: 2009-12-9 09:55
Previewing results对结果进行预览

Before you output results to your spreadsheet, you can preview the forecasted values and the best method selected with the Preview Forecast dialog.
在将结果输出在电子表各前，可以对预测值和最佳方法在Preview Forecast对话框中进行评估。

In the Preview Forecast dialog, you can preview the forecasted values for all the data series and all the methods run for each. After viewing the forecast values for each method, you can also override the selected method to use for the final forecasted values.
在Preview Forecast对话框中，你可以对所有方法和数据得出的预测值进行评估，在查看不同方法得到预测值后，可以得到最后的预测值。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.16.png (197.88 KB, 下载次数: 48)

Figure 1.16 Preview Forecast dialog

作者: lanj 时间: 2009-12-9 09:57
To continue with your tutorial:继续会案例进行操作：

20.Select Average Temperature from the Series list in the upper left corner of the Preview Forecast dialog.
20、选择平均温度查看预测值情况

Forecasted values appear for Average Temperature. Seasonal Additive is identified as the best-fitting method as in Figure 1.17.
在图1.17中选择最适合的预测方法后，平均气温的预测值显示出来

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.17.png (190.07 KB, 下载次数: 57)

Figure 1.17 Average temperature before Method override

作者: lanj 时间: 2009-12-9 09:58
21.Select Single Exponential Smoothing from the Method list.
The preview changes to show the forecast using single exponential smoothing instead of the seasonal additive method.
21、选择单边指数平滑方法
用单边指数平滑代替周期性递增的显示方法对预测结果进行预览

22.Click Override Best.
22、点击Override Best选择最佳情况。

This actually changes the forecast to use single exponential smoothing instead of the seasonal additive method as shown in Figure 1.18.
使用单一的预测指数平滑，而不是周期性递增的方法，如图1.18。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.18.png (203.07 KB, 下载次数: 54)

Figure 1.18 Average temperature after Method override

作者: lanj 时间: 2009-12-10 08:52
Summary概述
CB Predictor’s primary job is to create forecasts based on your historical data. The program’s method selections appear in the Preview dialog. When you override the program’s selections, you should carefully analyze your results.
CB Predictor最初是用来在历史数据的基础上进行预测。在Preview对话框中对方法进行选择，当给你选择了其中一种方法时，需要对结果仔细的分析。

In this example, there were three independent variables that combined to forecast the dependent variable, Usage. In the tutorial, you overrode the Average Temperature reading to use the method Single Exponential Smoothing instead of Seasonal Additive. What was the effect?
在这个案例中，有三个自变量会对因变量进行影响。在案例中平均温度因素选择单边指数平滑而不是季节附加，会有什么影响呢？

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.19.png (248.22 KB, 下载次数: 50)

作者: lanj 时间: 2009-12-10 08:54
本帖最后由 lanj 于 2009-12-10 08:57 编辑

Overriding the Average Temperature had a noticeable effect on the forecast (but not the fit) of the Usage variable.
平均温度对燃气的使用情况在预测表中有着明显的预测作用（但并不十分适合）。

23.Click Run.
23、点击Run
CB Predictor creates:

•The results pasted at the end of your historical data
在历史数据后面粘贴预测值。

•On a separate worksheet of your workbook, a report listing all the details of the time-series forecasts of each independent variable and the multiple linear regression of the dependent variable
在工作表中生成一个单独的sheet表，对时间序列预测的每个自变量和与因变量之间的多元线性回归生成详细的报告。

•On a separate worksheet of your workbook, a Methods Table page as PivotTables) listing all the time-series forecasting methods tried, their parameters, the error measures and statistics for each, and the regression parameters and statistics
在工作表中生成一个单独的sheet表，对时间序列预测的各种方法、参数和度量的偏差的统计数据、回归参数和统计数据生成一个方法列表。

24.Look at the results pasted below the historical data as shown in Figure 1.20.
24、图1.20中的历史数据后面粘贴的是预测的结果。
The pane was frozen beneath the column headers so they wouild appear in this figure.
被冻结的窗格下方的列标题显示了预测的数据

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.20.png (215.19 KB, 下载次数: 51)

作者: lanj 时间: 2009-12-10 08:57
本帖最后由 lanj 于 2009-12-10 09:11 编辑

Notice that the independent variables have been defined as Crystal Ball assumptions with normal distributions.
注意：自变量已经用水晶球假设的方法进行了定义，是正态分布。

25.Activate the Report worksheet and scroll to the section on the Average Temperature variable as shown in Figure 1.21.
25、如图生成报告，并查看平均温度的因素报告部分。

Notice the indication that the method used was an override of the best method.
注意：用覆盖的方法是最佳的方法。

26.Activate the Methods Table worksheet.
26、生成方法列表电子表。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.21.png (82.32 KB, 下载次数: 51)

图1.22.png (162.1 KB, 下载次数: 50)

作者: lanj 时间: 2009-12-11 09:03
本帖最后由 lanj 于 2009-12-11 09:05 编辑

27.Next to the Series button, select Average Temperature.
27、在序列按钮中，选择平均温度。

The table changes to show the parameters and statistics for each method of the Average Temperature forecast as shown in Figure 1.23.
在图1.23中就显示了平均温度的各种方法得到的相关参数和统计方法。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.23.png (167.59 KB, 下载次数: 66)

作者: lanj 时间: 2009-12-11 09:23
28.Click the Series button and drag it to the left of the Methods button.
28、点击序列按钮，并将其拖动到方法列表的左边。

The method table expands to include all the data series. When you drop the Series button next to the Methods button, the list of methods repeats for each series.
方法列表包括了所有的数据序列，当你把系列按钮拖动到方法列表左边后，方法列表显示的是每个序列的方法清单。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.24.png (166.17 KB, 下载次数: 61)

作者: lanj 时间: 2009-12-11 09:26
29.Click the down arrow beside the Table Items button.
A list of fields appears.
29、点击Table Items，显示其下拉清单

30.Uncheck all the items except for Rank.
30、除了排名以外全部不选择。

31.Click OK.
31、点击ok

The methods table changes to show one parameter: Rank. Look at the Average Temperature data. Under Rank, Single Exponential Smoothing is highlighted in blue, bold text to show that it was used to generate the results. Seasonal Additive, originally the best, is still listed with a Rank equal to 1.
方法列表显示其中一个参数的排名情况，查看平均温度数据，根据排名，单边指数平滑是最高的，用蓝色表示，粗体字表示过去一般情况下的结果。周期附加，最初最好的，是被列在排名第一的位置。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.25.png (165.71 KB, 下载次数: 53)

作者: lanj 时间: 2009-12-11 09:28
本帖最后由 lanj 于 2009-12-11 09:30 编辑

32.Move the Methods button to the left of the Series button.
32、将方法按钮移动到序列按钮的左边

The PivotTable reorganizes to show all the series grouped by method type as shown in Figure 1.26.
在图1.26中所有的序列以方法类别的方法进行显示，生成数据透析图。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图1.26.png (164.12 KB, 下载次数: 59)

作者: 漂在生活 时间: 2009-12-11 12:51
昨天没看呀，今天继续。come on

作者: lanj 时间: 2009-12-14 10:02
Chapter 2
Understanding the Terminology
理解相关术语
In this chapter在本章中
•Forecasting                                                 •Multiple linear regression
预测                                                             多元线性回归
•Time-series forecasting                               •Regression methods
时间序列预测                                                 回归方法
•Time-series forecasting error measures       •Regression statistics
时间序列预测误差度量                                        回归统计
•Time-series forecasting techniques             •Historical data statistics
时间序列预测技术                                           历史数据统计
•Time-series forecasting statistics
时间序列预测统计

This chapter describes forecasting terminology. It defines the time-series forecasting methods that CB Predictor uses, as well as other forecasting-related terminology.
本章对预测术语进行了描述，对CB Predictor工具中时间序列预测方法已经与预测相关的术语进行了定义。

This chapter also describes the statistics the program generates and the techniques that CB Predictor uses to do the calculations and select the best-fitting method.
本章同时也对统计数据生成的程序和方法进行了描述，CB Predictor使用这些计算公式选择最佳的方法。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-14 10:05
本帖最后由 lanj 于 2009-12-14 10:07 编辑

Forecasting预测
Forecasting refers to the act of predicting the future, usually for purposes of planning and managing resources. There are many scientific approaches to forecasting. You can perform “what-if” forecasting by creating a model and simulating outcomes, as with Crystal Ball, or by collecting data over a period of time and analyzing the trends and patterns. CB Predictor uses the latter concept, analyzing the patterns of a time series to forecast future data.
预测是指对未来行为的预测，通常的目的是进行规划和管理资源。有许多科学的预测方法。你可以使用CB Predictor方法通过建立模型和模拟结果来进行“假设”预测，或者通过收集一段时间的时间数据，分析其趋势和结构。CB Predictor使用最新的概念和分析方法对时间序列结构，从而预测未来的数据。

The scientific approaches to forecasting usually fall into one of several categories:
科学的预测方法通常分为以下几类：

Time-series  Performs time-series analysis on past patterns of data to forecast results. This works best for stable situations where conditions are expected to remain the same.
时间序列对以前数据用时间序列的方法分析结构来预测未来的结果。这个方法最适合用于预计未来将不会发生变化的稳定的情况。

Regression  Forecasts results using past relationships between a variable of interest and several other variables that might influence it. This works best for situations where you need to identify the different effects of different variables. This category includes multiple linear regression.
回归    根据因变量与会对其有影响的自变量因素之间的关系进行预测。这个方法最适合用于你需要确认不用因素分别有什么不同的作用的情
         况。回归方法包括多元线性回归。

Simulation  Randomly generates many different scenarios for a model to forecast the possible outcomes. This method works best where you might not have historical data, but you can build the model of your situation to analyze its behavior.
模拟    模拟是指在对未来可能的结果进行预测时会随机产生许多不同的情况形成一个模型。这种方法最适合用于你有可能没有历史数据但
         是你需要建立一个模型对事情的情况进行分析。

Qualitative  Uses subjective judgment and expert opinion to forecast results. These methods work best for situations for which there are no historical data or models available.
定性    通过主观和专家的观点对预测结果进行分析。这种方法最适合用于没有历史数据和模型的情况。

CB Predictor uses both time-series and multiple linear regression for forecasting. Crystal Ball uses simulation. Each technique and method has advantages and disadvantages for particular types of data, so often you might forecast your data using several methods and then select the method that yields the best results.
CB Predictor使用时间序列和多元线性回归的方法进行预测，水晶球使用模型。对于不通类型的数据。每种技术方法都有它的优点和缺点，所以一般情况下会使用多种方法来进行预测，并从中选择能产生最佳效果的其中一种方法

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-14 10:08
Time-series forecasting时间序列预测

Time-series forecasting assumes historical data is a combination of a pattern and some random error. Its goal is to isolate the pattern from the error by understanding the pattern’s level, trend, and seasonality. You can then measure the error using a statistical measurement to describe both how well a pattern reproduces historical data and to estimate how accurately it forecasts the data into the future. For more information on these error measurements, see “Time-series forecasting error measures” on page 42.
时间序列预测是假设历史数据是根据一定结构组合起来的，并且误差是随机的。它的目的是通过数据结构的水平、趋势和季节性将其中的误差寻找出来。你可以利用统计度量的方法对误差进行度量，从而对历史数据的结构进行复制，对未来预测数据的准确性进行描述。关于误差度量的更多的内容，可以看第42页的“时间序列预测误差度量”。

When you select different forecasting methods from the Methods Gallery, CB Predictor tries all of them. It then ranks them according to which method has the lowest error, depending on the error measure selected in the Advanced dialog. The method with the lowest error is the best method.
当你从概率分布图中选择不同的预测方法时，CB Predictor会对他们都进行尝试，然后根据Advanced对话框中误差度量选择的哪种方法的误差最小进行排名。误差最小的方法就是最好的方法。

There are two primary techniques of time-series forecasting used in CB Predictor. They are:
有两种主要时间序列预测的方法。它们是：

Nonseasonal smoothing  Estimates a trend by removing extreme data and reducing data randomness.
非季节性平滑       通过去除异常数据和降低数据随机性来对数据的趋势性进行评估的行为。

Seasonal smoothing Combines smoothing data with an adjustment for seasonal behavior.
季节性平滑          结合平滑的数据进行调整的季节性行为。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: 漂在生活 时间: 2009-12-14 12:11
{:9_328:} 继续停留，学习

作者: lanj 时间: 2009-12-15 17:03
今天差点忘记更新了，忙晕了{:8_314:}

作者: lanj 时间: 2009-12-15 17:04
Nonseasonal smoothing methods非季节性平滑

Smoothing models attempt to forecast by removing extreme changes in past data. The following methods are available.
平滑模型试图将以前数据中的异常点去除后进行预测，以下是各种方法：

Single moving average一次移动平均

Smooths out historical data by averaging the last several periods and projecting the last average value forward. CB Predictor can automatically calculate the optimal number of periods to average, or you can select the number of periods to average.
剔除过去几期的平均数据和预测的未来的值的异常数据，CB Predictor会自动计算出最佳的若干期的平均，或者你可以选择若干期的平均数据。

This method is best for volatile data with no trend or seasonality. It results in a straight, flat-line forecast.
这种方法最适合用于没有趋势和周期性的易变的数据，它的结果是直线、平线的预测。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.1.png (20.38 KB, 下载次数: 62)

作者: lanj 时间: 2009-12-15 17:05
Double moving average二次移动平均

Applies the moving average technique twice, once to the original data and then to the resulting single moving average data. This method then uses both sets of smoothed data to project forward. CB Predictor can automatically calculate the optimal number of periods to average, or you can select the number of periods to average.
使用两次移动平均技术，一次是对原始数据进行移动平均，第二次是对一次移动平均的数据的结果进行再一次的移动平均。这种方法是根据两次移动平均的数据进行评估。CB Predictor会自动计算出最佳的若干期的平均，或者你可以选择若干期的平均数据。

This method is best for historical data with a trend but no seasonality. It results in a straight, sloped-line forecast.
这种方法最适合用于有一定的趋势性但是没有周期性的历史数据。它的结果是直线、斜线的预测。

CB Predictor Note: For See Appendix B, “Time-Series Forecasting Method Formulas” for more information about the formulas CB Predictor uses for the following methods.
注意：附录B中“时间序列预测方法公式”，可以详细了解公式CB Predictor使用下列方法。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.2.png (19.1 KB, 下载次数: 60)

作者: lanj 时间: 2009-12-15 17:07
Single exponential smoothing (SES)单边指数平滑(SES）

Weights all of the past data with exponentially decreasing weights going into the past. In other words, usually the more recent data have greater weight. This largely overcomes the limitations of moving averages or percentage change models. CB Predictor can automatically calculate the optimal smoothing constant, or you can manually define the smoothing constant.
对以前的数据的权重用指数降低的方法进行加权，换句话说，就是越近期的数据的分量越重。这在很大程度上克服了移动平均和百分比变化模型的局限。CB Predictor可以自动计算平滑指数的最佳情况，或者你可以手动定义平滑指数。

This method is best for volatile data with no trend or seasonality. It results in a straight, flat-line forecast.
这种方法最适合用于有一定的趋势性但是没有周期性的历史数据。它的结果是直线、水平线的预测。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.3.png (19.85 KB, 下载次数: 56)

作者: lanj 时间: 2009-12-15 17:08
Holt’s double exponential smoothing (DES)霍尔特双边指数平滑（DES）

Double exponential smoothing applies SES twice, once to the original data and then to the resulting SES data. CB Predictor uses Holt’s method for double exponential smoothing, which can use a different parameter for the second application of the SES equation. CB Predictor can automatically calculate the optimal smoothing constants, or you can manually define the smoothing constants.
双边指数平滑是单边指数平滑进行两次，第一次是对原始数据进行指数平滑，第二次是对单边指数平滑的数据再进行指数平滑。CB Predictor在双边指数平滑时使用霍尔特方法，这种方法是第二次进行指数平滑时，与单边指数平滑公式的参数是不一样的。CB Predictor会自动计算最佳的平滑指数，或者你也可以选择自定义平滑指数。

This method is best for data with a trend but no seasonality. It results in a straight, sloped-line forecast.
这种方法最适合用于有趋势但是没有季节性的历史数据。它的结果是直线，斜线的预测。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.4.png (19.17 KB, 下载次数: 60)

作者: lanj 时间: 2009-12-15 17:09
Nonseasonal smoothing parameters非季节性平滑参数

There are several smoothing parameters used by the nonseasonal methods. For the moving average methods, the formulas use one parameter: period. When performing a moving average, you average over a number of periods. For single moving average, the number of periods can be any whole number between 1 and half the number of data points. For double moving average, the number of periods can be any whole number between 2 and one-third the number of data points.
使用非季节性方法有许多平滑参数。移动平均的方法，公式中有一个参数：周期。当你使用一个移动平均，你对一段时间的数据进行了平均。一次移动平均中，周期数可以是1到数据点一半数量的任何数。对于二次移动平均，周期数可以是2到数据点三分之一数量的任何数。

For single exponential smoothing, there is one parameter: alpha. Alpha (α) is the smoothing constant. The value of alpha can be any number between 0 and 1, not inclusive.
对单边指数平滑，有一个参数：平滑指数Alpha（α）。α的值可以是0到1之间除了0和1的的任何值。

For Holt’s double exponential smoothing, there are two parameters: alpha and beta. Alpha is the same smoothing constant as described above for single exponential smoothing. Beta (β) is also a smoothing constant exactly like alpha except that it is used during second smoothing. The value of beta can be any number between 0 and 1, not inclusive.
对于霍尔特双边指数平滑，有两个参数：alpha and beta，alpha和单边指数平滑描述的平滑指数一样， beta（β）也是和alpha一样的一个平滑指数，但是它是用在二次平滑中，这个值可以是0到1之间除了0和1的的任何值。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-15 17:11
明天说季节性平滑的相关方法

作者: 漂在生活 时间: 2009-12-15 17:35
我就等你~~~

作者: lanj 时间: 2009-12-16 09:02
Seasonal smoothing methods季节性平滑方法

Seasonal exponential smoothing methods extend the simple exponential smoothing methods by adding an additional component to capture the seasonal behavior of the data. There are four seasonal exponential smoothing methods used in CB Predictor.
季节性指数平滑方法是一个增加了额外的特点周期性行为的数据的单边指数平滑方法。在CB Predictor中一共有四种季节性指数平滑的方法。

Seasonal Additive Smoothing季节跌加平滑

Calculates a seasonal index for historical data that don’t have a trend. The method produces exponentially smoothed values for the level of the forecast and the seasonal adjustment to the forecast. The seasonal adjustment is added to the forecasted level, producing the Seasonal Additive forecast.
对没有趋势性的季节性特征的历史数据进行计算。这种方法得到的预测水平和季节性调整的预测值的指数平滑值。这种季节性调整是对预测水平的增强，得到季节跌加预测。

This method is best for data without trend but with seasonality that doesn’t increase over time. It results in a curved forecast that reproduces the seasonal changes in the data.
这种方法最适合用于没有趋势性但是季节性不顺着时间增加而增加的历史数据。它的结果是在数据复制季节性变化的弯曲预测。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.5.png (26.83 KB, 下载次数: 74)

作者: lanj 时间: 2009-12-16 09:04
Seasonal Multiplicative Smoothing季节乘积平滑

Calculates a seasonal index for historical data that don’t have a trend. The method produces exponentially smoothed values for the level of the forecast and the seasonal adjustment to the forecast. The seasonal adjustment is multiplied by the forecasted level, producing the Seasonal Multiplicative forecast.
对没有趋势性的季节性特征的历史数据进行计算。这种方法得到的预测水平和季节性调整的预测值的指数平滑值。种季节性调整是对预测水平的增强，得到季节乘积预测

This method is best for data without trend but with seasonality that increases or decreases over time. It results in a curved forecast that reproduces the seasonal changes in the data.
这种方法最适合用于没有趋势性但是季节性不顺着时间增加而增加的历史数据。它的结果是在数据复制季节性变化的弯曲预测。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.6.png (21.52 KB, 下载次数: 71)

作者: lanj 时间: 2009-12-16 09:05
Holt-Winters’ Additive Seasonal Smoothing霍尔特季节迭加平滑

Is an extension of Holt's exponential smoothing that captures seasonality. This method is based upon three equations that can be found in Appendix B. The method produces exponentially smoothed values for the level of the forecast, the trend of the forecast, and the seasonal adjustment to the forecast. This seasonal additive method adds the seasonality factor to the trended forecast, producing the Holt-Winters’ Additive forecast.
它是对有季节特征的霍尔特指数平滑的扩展。这种方法是在三个方程式基础上建立的，这三种方程式可以在附录B中查看。这种方法得到预测水平、预测趋势和季节性调整预测的指数平滑值。这种季节跌加方法增加了趋势预测季节性因素，得到霍尔特跌加预测。

This method is best for data with trend and seasonality that doesn’t increase over time. It results in a curved forecast that shows the seasonal changes in the data.
这种方法最适合用于有趋势性并且季节性不会随着时间增加而增加的数据。它的结果是弯曲预测，表示的是数据的季节性变化。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.7.png (20.62 KB, 下载次数: 61)

作者: lanj 时间: 2009-12-16 09:06
Holt-Winters’ Multiplicative Seasonal Smoothing霍尔特季节乘积平滑

Is similar to the Holt-Winters’ Additive method. This method also calculates exponentially smoothed values for level, trend, and seasonal adjustment to the forecast. This method's equations can also be found in Appendix B. This seasonal multiplicative method multiplies the trended forecast by the seasonality, producing the Holt-Winters’ Multiplicative forecast.
这种方法与霍尔特季节跌加方法很相似。它会对预测水平、趋势和季节性调整计算出指数平滑值。这种方法的公式可以在附录B中找到。这种季节乘积方法对预测季节性因素进行乘积，得到霍尔特季节乘积平滑。

This method is best for data with trend and with seasonality that increases over time. It results in a curved forecast that reproduces the seasonal changes in the data
这种方法最适合用于有趋势性并且季节性会随着时间增加而增加的数据。它的结果是在数据复制季节性变化的弯曲预测。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.8.png (20.32 KB, 下载次数: 68)

作者: lanj 时间: 2009-12-16 09:08
Seasonal smoothing parameters季节平滑参数

There are three smoothing parameters used by the seasonal methods: alpha, beta, and gamma.
有三个季节平滑参数：α，β，和γ。

alpha (α) Smoothing parameter for the level component of the forecast. The value of alpha can be any number between 0 and 1, not inclusive.
(α) 预测水平组成的平滑参数。这个值可以是0到1之间除了0和1的的任何值。

beta (β) Smoothing parameter for the trend component of the forecast. The value of beta can be any number between 0 and 1, not inclusive.
beta (β) 预测水平趋势的平滑参数。这个值可以是0到1之间除了0和1的的任何值。

gamma (γ) Smoothing parameter for the seasonality component of the forecast. The value of gamma can be any number between 0 and 1, not inclusive.
gamma (γ) 预测季节性的平滑参数。这个值可以是0到1之间除了0和1的的任何值。

Each seasonal method uses some or all of these parameters, depending on the forecasting method. For example, the seasonal additive smoothing method doesn’t account for trend, so it doesn’t use the beta parameter.
每种季节性方法根据预测方法的不通都有许多的参数。例如，季节跌加平滑方法不能计算趋势，所以它没有beta (β)参数。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-17 09:06
Time-series forecasting error measures时间序列预测误差度量

One component of every time-series forecast is the data’s random error that is not explained by the forecast formula or by the trend and seasonal patterns. The error is measured by fitting points for the time periods with historical data, and then comparing the fitted points to the historical data.
每个时间序列预测的一个组成部分是数据的随机偏差，这个偏差是无法通过预测公式和季节性结构趋势来解释的。偏差是对同一时间的历史数据和拟合点的度量的差异，并对它们进行比较。

All the examples are based on the set of data illustrated in the chart below. Most of the formulas refer to the actual points (Y) and the fitted points (^Y). In the chart below, the horizontal axis illustrates the time periods (t) and the vertical axis illustrates the data point values.
在下面的图中有案例说明，大多数的公式指的实际点是（Y），拟合点是（^Y），在下面的图表中横轴表示时间周期（t），纵轴表示数据点值。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.9.png (28.55 KB, 下载次数: 68)

作者: lanj 时间: 2009-12-17 09:07
CB Predictor measures the error using one of the following methods:
CB Predictor对误差度量使用下列方法之一：

•RMSE, below
•MAD, page 43
•MAPE, page 43
•均方根RMSE，如下
•平均绝对偏差MAD
•平均绝对百分比误差MAPE

RMSE
RMSE (root mean squared error) is an absolute error measure that squares the deviations to keep the positive and negative deviations from cancelling out each other. This measure also tends to exaggerate large errors, which can help eliminate methods with large errors.
RMSE（误差的平方根）是对绝对误差的度量，它是将误差进行平方对正负相互抵消。
这种方法也会对误差进行扩大，可以帮助我们通过扩大的误差进行消除。

MAD
MAD (mean absolute deviation) is an absolute error measure that originally became very popular (in the days before hand-held calculators) because it didn’t require the calculation of squares or square roots. While it is still fairly reliable and widely used, it is most accurate for normally distributed data.
MAD（平均绝对偏差）是对绝对误差的度量，在之前用手提电脑计算时非常受欢迎。因为它不需要计算平方或者平方根。现在它也十分可靠并被普通的使用，它能对正态分布数据计算的更精确。

MAPE
MAPE (mean absolute percentage error) is a relative error measure that uses absolute values. There are two advantages of this measure. First, the absolute values keep the positive and negative errors from cancelling out each other. Second, because relative errors don’t depend on the scale of the dependent variable, this measure lets you compare forecast accuracy between differently scaled time-series data.
MAPE（平均觉得百分比误差）是使用绝对值的相关误差的度量。这种度量有两个好处：第一是绝对值可以是的正负的误差相互抵消，第二由于相对误差不依赖于自变量的规模，这一措施可以对不同规模的时间序列数据之间更准确的预测。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-17 09:10
Time-series forecasting techniques时间序列预测技术

CB Predictor uses one of four forecasting techniques to perform time-series forecasting: standard, simple lead, weighted lead, and holdout.
CB Predictor使用时间序列预测四种预测技术中的一种：标准预测、简单领先预测、加权领先预测、抵抗预测。

Standard forecasting 标准预测

Standard forecasting optimizes the forecasting parameters to minimize the error measure between the fit values and the historical data for the same period. For example, if your historical data were:
通过减少同一周期的历史数据和拟合点之前的度量误差来优化标准预测的预测参数。

Period 1 2 3 4    5 6 7
Value 472 599  714  892 874  896  890
And your fit data were: 拟合值为：
Period 1 2 3 4    5 6 7
Value 488  609  702  888 890  909  870

CB Predictor calculates the RMSE using the differences between the historical data and the fit data from the same periods. For example:
CB Predictor对同一周期的历史数据和拟合点之间的差异用RMSE的方法进行计算。
(472-488)2 + (599-609)2 + (714-702)2 + (892-888)2 + ...

For standard forecasting, CB Predictor optimizes the forecasting parameters so that the RMSE calculated in this way is minimized.
对于标准预测，CB Predictor优化预测参数，使RMSE的计算方法最小化。

Simple lead forecasting简单领先预测

Simple lead forecasting optimizes the forecasting parameters to minimize the error measure between the historical data and the fit values, offset by a specified number of periods (lead). Use this forecasting technique when a forecast for some future time period has the greatest importance, more so than the forecasts for the previous or later periods.
通过减少同一周期的历史数据和拟合点之前的度量误差来优化简单领先预测的预测参数，从而抵消一定数量周期的情况。当对未来周期的预测远远比对近期或稍早时间的预测更为重要的时候使用这种预测技术。

For example, your company must order extremely expensive manufacturing components two months in advance, making any forecast for two months out the most important. For the same historical and fit example data described above in Standard Forecasting, CB Predictor calculates the RMSE using the difference between the historical data and the fit data from an offset number of periods (lead). With a lead of 2, the differences used in the RMSE calculation are:
例如，你的公司必须将花费极其巨大的制造部分提前2个月完成，对这两个月的预测是比其他更为重要的。同样是标准预测中案例的历史数据和拟合数据，CB Predictor使用RMSE的方法计算一段时期历史数据和拟合点之间的差异，如果要提前2个月，使用RMSE的方法就会不同：
(472-702)2 + (599-888)2 + (714-890)2 + (892-909)2 + ...

For simple lead forecasting, CB Predictor optimizes the forecasting parameters so that the RMSE calculated in this way is minimized.
对于简单领先预测，CB Predictor优化预测参数，使RMSE的计算方法最小化。

Weighted lead forecasting加权领先预测
Weighted lead forecasting optimizes the forecasting parameters to minimize the average error measure between the historical data and the fit values, offset by 0, 1, 2, etc., up to the specified number of periods (weighted lead). It uses the simple lead technique for several lead periods and then averages the forecast over the periods, optimizing this average value. Use this technique when the future forecast for several periods is most important.
加权导致预测预报优化参数，以尽量减少历史数据和拟合点之间的平均误差，从而抵消了0，1，2等，到指定的周期。，对前面的周期使用简单的领先技术，然后平均预测的周期，优化这一平均值。当预测未来几期是最重要的时候使用这种技术。

For example, your company must order extremely expensive manufacturing components zero, one, and two months in advance, making any forecast for all the time periods up to two months out the most important. For the same historical and fit example data described above in Standard Forecasting, CB Predictor calculates the RMSE for each lead up to the weighted lead using the difference between the historical data and the fit data from a set of offset periods (individual leads). With a weighted lead of 2, the differences used in the RMSE calculations are:
例如，您的公司必须事先以非常昂贵价格在开始的第1.2个月进行生产，对于两个月以内的任何时间段的预测是最重要的。对于相同的历史和上述标准预测中案例的数据，CB Predictor对一段时间内历史数据和适合数据之间的误差用加权的方法计算每个数值的RMSE。对领先2个周期的加权，在均方根误差计算中使用的区别是：
(472-488)2 + (599-609)2 + (714-702)2 + ... (lead of 0)
(472-609)2 + (599-702)2 + (714-888)2 + ... (lead of 1)
(472-702)2 + (599-888)2 + (714-890)2 + ... (lead of

Then CB Predictor averages the RMSE for the lead of 0, the lead of 1, and the lead of 2. For weighted lead forecasting, CB Predictor optimizes the forecasting parameters to minimize the average of the RMSE calculations.
CB Predictor会计算领先0个周期，领先1个周期，领先2个周期的RMSE平均值。CB Predictor预测的参数优化，以尽量减少对均方根误差计算的平均水平。

Holdout霍特尔预测
Holdout forecasting:

1.Removes the last few data points of your historical data.
1、去除过去一些数据点的历史数据

2.Calculates the fit and forecast points using the remaining historical data.
2、使用遗留下的历史数据计算拟合点和预测点

3.Compares the error between the forecasted points and their corresponding, excluded, historical data points.
3、对比预测数据和相对应已去除的数历史数据之间的误差。

4.Changes the parameters to minimize the error between the forecasted points and the excluded points.
4、改变参数，以减小预测点和已去除的点之间的误差。

CB Predictor determines the optimal forecast parameters using only the non-holdout set of data.
CB Predictor决定只使用没有去除的一系列数据来优化预测参数。

CB Predictor Note: If you have a small amount of data and want to use seasonal forecasting methods, using the holdout technique might restrict you to nonseasonal methods.
注意：如果你只有少部分的数据想用季节性预测方法，使用抵抗技术方法可以对非季节性方法进行控制。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-17 09:11
Time-series forecasting statistics时间序列预测统计

Theil’s U泰尔U预测

Theil’s U statistic is a relative error measure that compares the forecasted results with a naive forecast. It also squares the deviations to give more weight to large errors and to exaggerate errors, which can help eliminate methods with large errors. For the formula, see page 137.
泰尔U统计是相对误差的度量，它对预测结果和简单的预测进行比较。通常对误差进行平方，以便对最大和最小的误差进行加权，这样可以找到消除最大误差的方法。公式在137页。

Table 2.1 Interpreting Theil’s U

*Theil’s U statistic泰尔U统计***	*Means:*含义
Less than 1小于1	The forecasting technique is better than guessing.预测技术比预计的更好
1等于1	The forecasting technique is about as good as guessing.预测技术和预计的一样好
More than 1大于1	The forecasting technique is worse than guessing. 预测技术比预计的更差

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-17 09:13
明天说回归统计{:8_306:}

作者: 漂在生活 时间: 2009-12-17 10:13
我一直在等你更新。。。(weixiao:

作者: xinxijie531 时间: 2009-12-17 10:24
感谢分享！继续努力哈！

作者: 米老鼠 时间: 2009-12-17 10:36
非常有帮助的文章,顶一个!

作者: lqbz 时间: 2009-12-17 13:14
好好学习，天天向上。

作者: lanj 时间: 2009-12-18 10:15
Multiple linear regression多元线性回归

Multiple linear regression is used for data where one data series (the dependent variable) is a function of, or depends on, other data series (the independent variables). For example, the yield of a lettuce crop depends on the amount of water provided, the hours of sunlight each day, and the amount of fertilizer used.
多元线性回归方法一般是用在一系列数据（因变量）是有影响的，或浇水的数量，每天日照的小时和每天的化肥使用数量。

The goal of multiple linear regression is to find an equation that most closely matches the historical data. The word “multiple” indicates that you can use more than one independent variable to define your dependent variable in the regression equation. The word “linear” indicates that the regression equation is a linear equation.
多元线性回归的目的是找到最适合历史数据的方程式。“多元”是指在回归方程有一个以上的自变量会对因变量有影响。“线性”是指回归方程是一个线性方程。

The linear equation describes how the independent variables (x1, x2, x3,...) combine to define the single dependent variable (y). Multiple linear regression finds the coefficients for the equation:
线性方程是指有多个自变量（x1、x2、x3…）如何对单独的一个因变量（y）的关系。
y = b0 + b1x1 + b2x2 + b3x3 + ... + e

where b1, b2, and b3, are the coefficients of the independent variables, b0 is the y-intercept, and e is the error.
b1、b2和b3是自变量的系数，b0是y的截距，e是误差。

If there is only one independent variable, the equation defines a straight line. This uses a special case of multiple linear regression called simple linear regression, with the equation:
如果只有一个自变量，方程得到的是一条直线。这种情况的多元线性回归成为一元线性回归，方程是：
y = b0 + b1x + e

where b0 is where on the graph the line crosses the y axis, x is the independent variable, and e is the error. When the regression equation has only two independent variables, it defines a plane. When the regression equation has more than two independent variables, it defines a hyperplane.
b0是图表中图线与y轴相交的值，x是自变量，e是误差。当回归方程只有两个自变量时，形成的是一个平面。当回归方程有2个以上的自变量时，形成的是一个超平面。

To find the coefficients of these equations, CB Predictor uses singular value decomposition. For more information on this technique, see “Singular value decomposition” on page 143.
为了找到这些方程的系数，CB Predictor会使用奇异值分解的方法。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图2.10.png (50.03 KB, 下载次数: 62)

作者: lanj 时间: 2009-12-18 10:17
下面说回归方法，通常用的有三种方法

作者: lanj 时间: 2009-12-18 10:18
本帖最后由 lanj 于 2009-12-18 10:19 编辑

Regression methods回归方法

CB Predictor uses one of three regression methods: standard regression, forward stepwise regression, and iterative stepwise regression.
CB Predictor使用三种回归方法中的一种：标准回归，逐步回归和迭代回归。

Standard regression标准回归

Standard regression performs multiple linear regression, generating regression coefficients for each independent variable you specify, no matter how significant.
标准回归是多元线性回归，根据要求对每个自变量生成回归系数，各系数没有轻重之分。

Forward stepwise regression逐步向前回归

Forward stepwise regression adds one independent variable at a time to the multiple linear regression equation, starting with the independent variable with the most significant probability of the correlation (partial F statistic). It then recalculates the partial F statistic for the remaining independent variables, taking the existing regression equation into consideration.
逐步回归是指在某个时间在多元线性方程中增加一个自变量，首先从相关性最大、最重要的自变量开始（部分F统计）。然后重新计算增加了自变量的部分f统计，同时考虑现有的回归方程。

CB Predictor Note: The resulting multiple linear regression equation will always have at least one independent variable.
注意：多元线性方程的至少要有一个以上自变量。

Forward stepwise regression continues to add independent variables until either:
•It runs out of independent variables.
•It reaches one of the selected stopping criteria in the Stepwise Options dialog.
•The number of included independent variables reaches one-third the number of data points in the series.
逐步回归中连续增加自变量直到：
自变量全部增加完了
根据逐步回归方法达到了停止的标准
在系列中自变量的数量达到了1/3的数据点的数量。
There are two stopping criteria: 有两个停止标准：

R-squaredR平方

Stops the stepwise regression if the difference between a specified statistic (either R2 or adjusted R2) for the previous and new regression solutions is below a threshold value. When this happens, CB Predictor does not use the last independent variable.如果之前的统计和新的回归方法之间的偏差低于临界值的时候应该停止逐步回归。当这种情况发生的时候，CB Predictor不使用最后一个自变量。For example, the third step of a stepwise regression results in an R2 value of 0.81, and the fourth step adds another independent variable and results in an R2 value is 0.83. The difference between the R2 values is 0.02.If the Threshold value is 0.03, CB Predictor returns to the regression equation for the third step and stops the stepwise Regression例如，逐步回归结果的第三步R2的值是0.81，第四步加另一个独立变量后R2的值是0.83。两者之间的R2值差值是0.0 2。
如果差值为0.03，CB Predictor将返回到第三步回归方程，并停止逐步回归。

F-test significanceF值

Stops the stepwise regression if the probability of the F statistic for a new solution is above a maximum value.如果新方案的F值是最大的则停止逐步回归。For example, if you set the maximum probability to 0.05 and the F statistic for the fourth step of a stepwise regression results in a probability of 0.08, CB Predictor returns to the regression equation for the third step and stops the stepwise regression.例如，如果设置的最大概率值是0.05，逐步回归结果第四步F统计值是0.08，CB Predictor返回到第三步回归方程，并停止逐步回归

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-18 10:21
Iterative stepwise regression迭代回归

Iterative stepwise regression adds or removes one independent variable at a time to or from the multiple linear regression equation.
迭代回归是指在某个时间点从多元线性回归方程中增加或去除一个自变量。
To perform iterative stepwise regression, CB Predictor:迭代回归的情况是：

1.Calculates the partial F statistic for each independent variable.
1、对每个自变量计算其部分F值。

2.Adds the independent variable with the most significant correlation (partial F statistic).
2、增加相关性最明显的自变量（部分F值）

3.Checks the partial F statistic of the independent variables in the regression equation to see if any became insignificant (have a probability below the minimum) with the addition of the latest independent variable.
3、检查回归方程中的每个自变量的部分F值，看它们是否会因为增加了一个自变量而变的没有相关性（低于最小值）。

4.Removes the least significant of any insignificant independent variables one at a time.
4、在某个时间点去除相关性最小的自变量。

5.Repeat step 3 until no insignificant variables remain in the regression equation.
5、重复第三步，直到回归方程中没有没有相关性的自变量。

6.Repeat steps 1 through 5 until:
•The model runs out of independent variables.
•The regression reaches one of the stopping criteria (see the previous section for information on how the stopping criteria work).
•The same independent variable is added and then removed
6、重复第一到第五步，直到：
 使用完了所有的自变量
 达到了停止回归的标准
 同一个进行了自变量增加、删除。

CB Predictor Note: The resulting equation will always have at least one independent variable.
CB Predictor注意：方程最后至少要包含一个自变量

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-18 10:22
回归统计涉及了几个统计参数，如R方等.....

作者: lanj 时间: 2009-12-18 10:23
本帖最后由 lanj 于 2009-12-18 10:24 编辑

Regression statistics回归统计

Once CB Predictor finds the regression equation, it calculates several statistics to help you evaluate the regression. See Appendix E, “Regression Statistic Formulas” for more information on the formulas CB Predictor uses to calculate these statistics.
一旦CB Predictor得到回归方程，它会帮助你计算各种统计数据来对回归进行评。附录E中“回归统计公式”有关于CB Predictor计算统计的更多内容。

R2

Coefficient of determination. This statistic indicates the percentage of the variability of the dependent variable that the regression equation explains.
R2是一个决定系数。它能表示在回归方程中因变量受到多大影响的百分比的情况。

For example, an R2 of 0.36 indicates that the regression equation accounts for 36% of the variability of the dependent variable.
例如，在回归方程中R2的值是0.36的话，就意味者自变量对因变量有36%的影响力。

Adjusted R2调整 R2

Corrects R2 to account for the degrees of freedom in the data. In other words, the more data points you have, the more universal the regression equation is. However, if you have only the same number of data points as variables, the R2 might appear deceivingly high. This statistic corrects for that.

调整后的R2是在数据自由度方面对R2进行修改，换句话说，你如果有更多的数据点，回归方程就越广泛。但是，如果你只有相同的数据点，R2会出现很高的假象。这种统计方法可以对此进行纠正。

For example, the R2 for one equation might be very high, indicating that the equation accounted for almost all the error in the data. However, this value might be inflated if the number of data points was insufficient to calculate a universal regression equation.

例如，在一个方程中的R2可能非常高，这表明方程能计算出几乎所有的误差值。但是如果数据点的数量不足以用来进行回归方程计算时，这个值也可能非常大。

SSE

Sum of square deviations. The least squares technique for estimating regression coefficients minimizes this statistic, which measures the error not eliminated by the regression line.
平方差的总和。用最小的计算方法来估计回归系数的最小值，从而来度量回归线中没有去除的偏差。

For any line drawn through a scatter plot of data, there are a number of different ways to determine which line fits the data best. One method is to compare the fit of lines is to calculate the SSE (sum of the squared errors) for each line. The lower the SSE, the better the fit of the line to the data.
任何图线都可以通过散点图形成，但是有一些不同的方法去决定那条直线表明最好的数据。其中一中比较线性好坏的方法就是SSE。SSE越低，表示图线和数据越适合。

F statistic

Tests the significance of the regression equation as measured by R2. If this value is significant, it means that the regression equation does account for some of the variability of the dependent variable.
通过度量R2来测试回归方程的影响力。如果这个值关联性大，则表示回归方程中自变量的可变性越大。

t statistic

Tests the significance of the relationship between the coefficients of the dependent variable and the individual independent variable, in the presence of the other independent variables. If this value is significant, it means that the independent variable does contribute to the dependent variable.
测试自变量系数和单独的因变量或其他的因变量之间关系的重要性。如果这个值是重要的，那表示因变量受自变量的影响大。

p

Indicates the probability of your calculated F or t statistic being as large as it is (or larger) by chance. A low p value is good and means that the F statistic is not coincidental and, therefore, is significant. A significant F statistic means that the relationship between the dependent variable and the combination of independent variables is significant.
表示计算F的可能性或者统计的数据越来越大。P值越低表示越好，并意味着F统计的可能性越高，因此更有意义。一个有效的F值表示自变量和相关的因变量之间的关系越明显。

Generally, you want your p to be less than 0.05.
一般来说，P值最好小于0.05

Durbin-Watson

Detects autocorrelation at lag 1. This means that each time-series value influences the next value. This is the most common type of autocorrelation.For the formula, see page 138.
The value of this statistic can be any value between 0 and 4. Values indicate slow-moving, none, or fast-moving autocorrelation, as shown in Table 2.2.
缺陷自相关是1。这表示每个时间序列值会对下一个值有影响。这是一般情况下的自相关。这种统计的值可以是0-4之间的任何数值，数值表示移动快慢的情况。

Table 2.2 Interpreting the Durbin Watson statistic

*Durbin-Watson statistic*统计	*Means*含义
Less than 1 小于1	The errors are positively correlated. An increase in one period follows an increase in the previous period.误差是正相关。一个周期增加下一个周期也会随着增加。
22	No autocorrelation.不相关
More than 3大于3	The errors are negatively correlated. An increase in one period follows an decrease in the previous period.误差是负相关。一个周期的增加下一个周期会随之减少。

Avoid using independent variables that have errors with a strong positive or negative correlation, since this can lead to an incorrect forecast for the dependent variable.

应该避免自变量有明显的正相关或者负相关，这样会导致因变量预测不准确。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-18 10:33
本帖最后由 lanj 于 2009-12-22 13:25 编辑

Historical data statistics历史数据统计

CB Predictor automatically calculates the following statistics for historical data series:
CB Predictor会自动计算历史数据的下列统计值：

mean均值
The mean of a set of values is found by adding the values and dividing their sum by the number of values. The term “average” usually refers to the mean. For example, 5.2 is the mean or average of 1, 3, 6, 7, and 9.
平均值是指一系列数据通过增加数据或者减少数据的和，通常称为平均值。例如5.2就是数据1.3.6.7.9的平均值。

standard deviation标准方差
The standard deviation is the square root of the variance for a distribution. Like the variance, it is a measure of dispersion about the mean and is useful for describing the “average” deviation.
标准方差是一个分布的偏差的平方根。

For example, you can calculate the standard deviation of the values 1, 3, 6, 7, and 9 by finding the square root of the variance that is calculated in the variance example below.
The standard deviation, denoted as , is calculated from the variance as follows:
例如，你可以对1.3.6.7.9这组数字计算它们的标准方程，在下面的例子中可以看出如何用方差计算方差的平方根。
标准偏差，计算如下：
S=√10.2＝3.19

where the variance is a measure of the dispersion, or spread, of a set of values about the mean. When values are close to the mean, the variance is small. When values are widely scattered about the mean, the variance is larger.
偏差是对一系列数值的平均值进行分散，或离散计算。当这个值接近平均值，差异很小。当这个值越远离平均值，差异越大。

To calculate the variance of a set of values:
计算数值的误差值：

1.Find the mean or average.
1、找到数据的平均值。

2.For each value, calculate the difference between the value and the mean.
2、对每一个值，计算其和平均值之间的差值。

3.Square these differences.
3、对这些差值进行平均。

4.Divide by n-1, where n is the number of differences.
4、除以n-1，n是表示有多少个数据。

For example, suppose your values are 1, 3, 6, 7, and 9. The mean is 5.2. The variance, denoted by , is calculated as follows:
例如，1.3.6.7.9这组数据，它们的均值是5.2，方差的计算方法如下：

S^2=[(1-5.2)^2+(3-5.2)^2+(6-5.2)^2+(7-5.2)^2+(9-5.2)^2]/(5-1)=40.8/4=10.2

minimum最小值
The minimum is the smallest value in the data range.
最小值是指在一组数据中最小的一个数值。

Maximum最大值
The maximum is the largest value in the data range.
最大值是指在一组数据中最大的一个数值。

Ljung-Box statistic Ljung-Box 统计
Measures whether a set of autocorrelations is significantly different from a set of autocorrelations that are all zero. See page 139 for the formula.
对自相关是0和自相关明显的差异进行度量。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-18 10:35
下次进入手册的第三章，用CB Predictor进行预测{:8_307:}

作者: 思步 时间: 2009-12-18 13:21
:dabin期待ing

作者: lanj 时间: 2009-12-21 09:02
Chapter 3
Forecasting with CB Predictor
用CB Predictor进行预测

In this chapter
•Overview
•Creating a spreadsheet with historical data
•Loading and starting CB Predictor
•Guidelines for using CB Predictor
•Analyzing the results
•Customizing reports and charts
在这一章中
•概述
•用历史数据创建一个电子表格
•加载和启动CB Predictor
•使用CB Predictor指南
•分析结果
•自定义报告和图表

This chapter contains detailed procedures for using CB Predictor. It describes how to forecast using both time-series forecasting methods and multiple linear regression. It also describes all the settings you can choose for your results.
本章对CB Predictor的使用步骤进行了详细描述。它对时间序列和多元线性回归的预测都进行了描述，同时对选择结果的设置进行了描述。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-21 09:03
Overview概述

When you use CB Predictor for the first time, there are several things you need to do:
当你第一次使用CB Predictor时，需要注意以下几方面的事情：

1.Create an Excel spreadsheet with your historical data.
1、建立一个有历史数据的excel电子表格

2.Start CB Predictor.
2、打开CB Predictor

3.Run CB Predictor.
3、运行CB Predictor

4.Analyze your results.
4、对结果进行分析

5.Customize your results.
5、对结果进行设置。

This chapter describes how to complete each of these steps to make forecasting your historical data an easy task.
本章描述了怎样完成对历史数据进行预测的每一个步骤。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-21 09:04
本帖最后由 lanj 于 2009-12-21 09:07 编辑

Creating a spreadsheet with historical data
用历史数据创建一个电子表格

Before using CB Predictor, you must create an Excel spreadsheet with your historical data. Creating a spreadsheet for use with CB Predictor is easy. The spreadsheet must include:
在使用CB Predictor之前，你必须先建立一张有历史数据的excel电子表格。这张电子表格必须包括：

•Optionally, a descriptive spreadsheet title.
对电子表格标题进行描述。

•Optionally, a date (or other time period, such as Q2-2004) column or row, either at the top or along the left side of your data. If you format your dates as Excel dates, CB Predictor’s Intelligent Input can find the dates, extend them with the forecasted values, and use them as labels on charts.
有一组数据按列或者行进行排列，如果对数据继续排列，CB Predictor会自动的选择输入的数据，并且会输出预测值。

•Historical data, spaced equal time periods apart, in columns or rows adjacent to the date column or row. You can use CB Predictor to simultaneously forecast from one to 10,000 adjacent historical data series.
历史数据必须是同一时间段的数据，按列或者行排列，你可以使用CB Predictor同时对相近的1-10000的历史数据进行预测。

Excel Note: Excel only has 256 possible columns, compared to 16,000 or 65,000 possible rows (depending on your version). So, if you have a large number of historical data points, organize your data in columns. If you have a large number of data series, organize your data in rows.
注意：EXCEL表只有256列，16000或65000行（根据软件的不同版本）。所以，如果你的历史数据比较多，需要对历史数据进行整理。

To produce a reasonable forecast, you should have at least 6 historical data values. To use seasonal forecasting methods, you need at least two complete cycles of data.
为了得到一个合理的预测值，你需要至少有6个以上的历史数据。如果使用季节性预测方法，你不要至少有两个以上完整周期的数据。

•Optionally, headings for each data column or row, such as SKU 23442, Gas Usage, or Interest Rate.
对每行或每列的数据输入标题。

The Toledo Gas spreadsheet has all these components.
托莱多气体电子表格的所有内容：

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图3.1.png (191 KB, 下载次数: 69)

作者: lanj 时间: 2009-12-21 09:08
Loading and starting CB Predictor
加载和启动CB Predictor

If you have Crystal Ball Professional or Premium Edition, CB Predictor is loaded when you start Crystal Ball.

如果你有CB Predictor的专业版或高级版，CB Predictor会在你启动Crystal Ball的时候同时启动
To start CB Predictor, use one of these:

•In the menubar, choose Run > CB Predictor, or
•Type Alt-r, p.

When you start CB Predictor, the CB Predictor dialog, or wizard, appears as shown in Figure 3.2. It has four tabs, arranged from left to right in the typical order of their use. For descriptions of each setting on each tab, see Appendix A, “The CB Predictor Wizard.”
开始CB Predictor，在菜单中点击Run>CB Predictor，或者点击Alt-r，p
当开始CB Predictor，CB Predictor对话框或者向导会出现，如图3.2.从左到右它有四个表格。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-21 09:10
本帖最后由 lanj 于 2009-12-21 09:15 编辑

Guidelines for using CB Predictor CB Predictor使用指南

After you create your Excel spreadsheet with historical data, forecasting data using CB Predictor follows a 10-step process:
将历史数据放入到Excel电子表格中后，按照以下的10个步骤用CB Predictor对历史数据进行预测：

1.Select a cell range with your historical data to forecast.
1、选择历史数据中的其中一个单元格来进行预测。

2.Specify your data arrangement.
2、指定数据的排列方式，

3.View a graph of your historical data to identify any seasonality (data cycles) or trend and to see summary statistics.
3、查看历史数据确认数据的季节性或者趋势和概要统计。

4.Identify your time periods, whether your data have seasonality, and, if so, how long your season is.
4、确认时间周期，看数据是否有季节性，如果有，周期是多长。

5.Set whether you want to use multiple linear regression to forecast any variables.
5、设定对预测变量是否需要使用多元线性回归方法。

6.Select the time-series forecast methods to try for each variable.
6、对每一个因素选择时间序列预测方法

7.Enter the number of periods you want to forecast.
7、输入你想要预测的周期数

8.Select a confidence interval to calculate or display with your forecasted values.
8、选择一个置信区间来计算或者显示预测值。

9.Select the results you want.
9、选择你需要的结果

10.Preview and run the forecast, creating your results.
10、评审和运行预测，得到你的结果。

CB Predictor leads you through these steps using the CB Predictor wizard, but you can also go directly to any of these steps to change methods or select different settings and reforecast data.
CB Predictor使用向导引导每个步骤，但是你也可以直接到其中的任何一个步骤去选择不同的预测结果。

Also, many of these steps might be done automatically by CB Predictor. For example, CB Predictor’s Intelligent Input can find and select your data, their arrangement, and the data units. And, you can skip many other steps if the settings are already what you want. For example, all the time-series forecasting methods are selected by default, and that is probably how most users will leave them.
同时，许多步骤CB Predictor会自动执行，例如，CB Predictor自动选择你要输入的数据，数据排列，和数据单位。你可以通过设置你想要的跳过一些步骤。

Selecting historical data选择历史数据

When you select your historical data, you must identify the Excel cells that contain the data, including dates and headers, and set settings to identify dates, headers, and orientation in the data.
当你选择了历史数据，你必须确认excel单元格包括了这些数据，包括数值和标题等。

To produce a reasonable forecast, you should have a minimum number of historical data points. CB Predictor imposes several requirements defining a minimum number of points to use to create a forecast. The bare minimum is 5. However, other limitations include:
为了得到一个合理的结果，你需要有历史数据的最小值。CB Predictor

•Single moving average requires that the number of historical data points is twice the number of points to forecast.
一次移动平均要求历史数据的数量是需要预测的数据点数量的两倍。

•Double moving average requires that the number of historical data points be three times the number of points to forecast (or at least 6, whichever is higher).
二次移动平均要求历史数据的数量是需要预测的数据点数量的三倍（或者至少是6个）

•To use seasonal methods, you must have at least two seasons (complete cycles) of historical data.
使用季节性方法，你必须有至少两个季节（完整周期）的历史数据

•For multiple linear regression, the number of historical data points must be three times the number of independent variables (counting the included constant as an independent variable).
对于多元线性回归，历史数据点的数量必须是自变量数量的三倍以上

•To lag an independent variable in multiple linear regression, the number of historical data points must be three times the lag.
在多元线性回归中要对一个自变量进行滞后，历史数据的数量必须是滞后次数的三倍。

If your data has empty cells in the middle of a data series, CB Predictor returns an error. CB Predictor treats zeros in data series as data values. If you are trying to forecast several data series at once, your data series do not have to start at the same time period. However, all the data series must end at the same time period.
如果你的数据系列中有空的单元格，CB Predictor会提示错误。CB Predictor会把它默认为是0。如果你想一次预测更多的数据，你的数据系列中不需要都是同一个时间开始的数据，但是需要是在同一个时间结束的数据。

When you initially open a spreadsheet, there are three ways to select the historical data to forecast:
当你开始打开一个电子表格时，有三种方法去选择对历史数据进行预测：

•Use CB Predictor’s Intelligent Input.
使用CB Predictor自动选择

•Select your data before you start the wizard.
在开始向导之前选择你的数据

•Select your data after you start the wizard.
在选择数据后再开启向导

Automatic data selection数据自动选取

The easiest way to select data is to select one cell somewhere in your continuous data range before you start the CB Predictor wizard. When you start the wizard, CB Predictor’s Intelligent Input searches for all the adjacent cells with numbers, dates, and headers and makes some other assumptions about your input data, such as whether your data are in rows or columns. This often completes most of the fields and settings on the Input Data tab.
这是最简单的方法，在打开CB Predictor向导前在数据中选择其中的一个数据。当你打开向导，CB Predictor自动选择所有数据的单元格，包括数据、标题和对数据的说明。

Manual data selection before starting CB Predictor

在用 CB Predictor之前选择数据的方法

The second way to select your historical data is to highlight the data range (including headers and dates) before you start CB Predictor.
第二种方法是在开始CB Predictor之前选择历史数据（包括标题和数据）

Manual data selection within CB Predictor用 CB Predictor选择数据的方法

The third way to select historical data is to start CB Predictor with no data, date, or header cells selected. The Input Data tab of the wizard appears with the Range field blank. At this point, you must select your historical data manually.
第三种方法是先CB Predictor，然后选择历史数据。

To select historical data manually from the Input Data tab:
在数据输入表中选择历史数据的方法是：

1.Start CB Predictor.
The Input Data tab of the wizard dialog appears. For more information on this dialog, see “Input Data tab” on page 100.
1、开始CB Predictor，输入数据的向导对话框出现。

2.Under Step 1, in the Range field, either type a range name (if defined), enter the range of cells with the historical data, including any headers (e.g., A4:B42), or:
2、第一步之后，在范围表格中，选择历史上历史数据，包括标题等（如，A4:B2），或者

a. Click Select.
The Select Range dialog replaces the wizard dialog.
a．点击select，出现选择数据范围的对话框

b. Select the cells with the historical data, including any cells with headers and dates.
b，选择历史数据的单元格，包括所有的标题和数据。

c. Click OK to return to the wizard dialog.
The selected range appears in the Range field.
c、点击ok，选择的数据范围出现在范围框中。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-21 09:16
本帖最后由 lanj 于 2009-12-21 09:18 编辑

Specifying data arrangement对数据进行排列

No matter which selection method you use, you must also specify your data arrangement to help CB Predictor identify whether you have dates and headers adjacent to your data series and whether your data are in rows or columns. If you used the Intelligent Input to select your data, these settings should already be set correctly.
无论你使用那种选择数据的方法，你还必须指定你的数据的安排，以帮助CB预测确定数据序列是否有日期和标题，以及是否是按要求进行行或列排列。如果使用数据是用智能输入的，这些设置都应该已经正确设置。

To set arrangement settings, use the Input Data tab of the CB Predictor wizard as shown in Figure 3.2 on page 64
排列设置，参考第64页的表3.2 CB预测显示卡的方法输入数据：

1.Under Step 2 of the wizard, if your historical data is in:
•Rows, select Data In Rows
•Columns, select Data In Columns
1、如果历史数据是如下情况，根据向导的第二步：
行，选择数据以行排列
列，选择数据以列排列

2.If you have headers (titles) at the top of your columns or to the left of your rows, select the First Row (Column) Has Headers setting.
2、如果你在每列的开头或者每行的左边都有标题，选择第一行（列）作为标题

3.If your first column or row lists the dates or time periods for your data series, select the First Column (Row) Has Dates setting.
3、如果你的第一行或列的时间有日期或者时间周期，则选择第一行（列）作为日期设置。

Viewing your historical data查看历史数据

As you progress through the wizard, you need to know if your data are seasonal (increase and decrease in a regular cycle) and, if so, what the season or cycle is. If you don’t already have a feel for the behavior of your data, you might want to view your selected historical data before you continue.
在使用向导卡的时候，需要知道数据是否是有周期性的（增加或者减少是有一定的规律的），如果是，是按怎样的周期或者规律。如果在之前对数据的情况不是很了解，你可能要在下一步操作之前先查看你的数据。

To view a graph of your historical data: 查看图表中的历史数据：

1.Under Step 3 of the CB Predictor wizard (on the Input Data tab), click View Data.
1、在CB Predictor的第三步向导卡（输入数据表格），点击查看数据

The View Historical Data dialog appears as shown in Figure 3.3.
如图3.3，查看历史数据对话框会显示出来。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图3.3.png (207.7 KB, 下载次数: 68)

作者: lanj 时间: 2009-12-21 09:21
本帖最后由 lanj 于 2009-12-21 09:24 编辑

2.If your data have a recurring pattern, your data might be seasonal, and you should note how many dates or time periods are in one cycle of the pattern.
2、如果你的数据有反复出现的情况，你的数据可能存在周期性，你应该注意到有多少日期或时间段是在一个周期的规律。

3.If you selected more than one historical data series, change the graph to view another data series by selecting it from the Series list.
3、如果选择了一个或更多的历史数据序列，从序列列表中选择要查看其他序列的数据。

CB Predictor Note: When the Series list is selected, you can also use the up and down arrows to scroll through the list.
CB Predictor 注意：但选择序列数据时，可通过列表的上下滑轮来进行选择。

4.To see the three highest autocorrelations and the Ljung-Box statistic, select Autocorrelations from the View list.
4、看三个最高的自相关和Ljung-Box统计，从查看列表中选择自相关。

The View Historical Data dialog appears in Autocorrelations view as shown in Figure 3.4.
如图3.4.历史数据对话框胡自动显示出来：

For information on the Ljung-Box statistic, see “Ljung-Box statistic” on page 53.
更多关于Ljung-Box 统计的信息，查看 “Ljung-Box统计”
For more information on both views of the View Historical Data dialog, see page 102.
更多关于历史数据对话框的信息，可以查看手册第102页

5.Click Close.
The Input Data tab reappears.
5.点击关闭
数据输入表显示出来。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图3.4.png (185.73 KB, 下载次数: 58)

作者: lanj 时间: 2009-12-21 09:39
本帖最后由 lanj 于 2009-12-21 09:53 编辑

Identifying time periods and seasonality确认时间周期和季节周期

You need to identify your time periods and seasonality for CB Predictor.
To identify your data’s time periods and seasonality:
你需要确认CB Predictor中时间周期。

1.In the Input Data tab, click Next.
The Data Attributes tab appears. For more information on this tab, see “Data Attributes tab”
1、在输入数据表格中，点击下一步
数据属性在表格中显示，如果需要查看表格的更多信息，点击：数据属性表

2.Under Step 4 on the Data Attributes tab, identify the time period for your data values.
For example, if your data represent monthly numbers, select months.
2、在数据属性表的第四步，确认数据的时间周期。
例如，如果你的数据是按月份排列，选择月份

3.Indicate the seasonality of your data:
•If any of your data series are seasonal, select the Seasonality setting and enter the number of time periods it takes before your data pattern repeats. You must have at least two seasons (complete cycles) of data to use the seasonal methods.
3、标明数据的周期性
如果你的数据有周期性，选择Seasonality并输入重复的周期是多少。如果使用周期性的方法，你必须有至少两个完整周期的数据。

This number is usually the number of periods per year. For example, if you have 24 monthly data points, and your data has peaks every December, your seasonality (repeating pattern) has a period of one year or 12 months.
这个数据通常是以每年为周期的，例如，如果你有24个月的数据，而且是以12个月为一个周期，那数据的周期性就是一年或者12个月为一个周期。

CB Predictor Note: You can also view the autocorrelations on the View Historical Data dialog to discover how many periods you have in a season.
•If none of your data are seasonal, select the No Seasonality setting.
•If you are forecasting multiple data series and each has a different seasonality, you must forecast each individually.
注意：你也可以通过查看历史数据图表中的自相关来发现数据有多少个周期。
       如果数据没有周期性的话，选择No Seasonality
       如果要预测多组数据，并且每组数据有不同的周期性，你必须对单个进行预测。

Using multiple linear regression使用多元线性回归

If you know that some independent variables affect another variable of interest, you should use multiple linear regression as the forecasting method for that particular dependent variable. For example, summer temperatures affect electricity usage because as it gets hotter, more people run their air conditioning. This means that electricity usage (the dependent variable) is dependent on the temperature  （  an independent variable).
如果你发现有些自变量会对其他变量有影响，那在对特定的因变量进选择预测方法的时候，你需要使用多元线性回归的方法。
例如，夏天的气温会因为变高，更多的人使用空调等原因，影响电量的使用情况。这就表明电量的使用情况的多少（因变量）是依赖于温度（一个自变量）的影响的。

CB Predictor Note: The “multiple” in multiple linear regression represents the fact that you can have more than one independent variable.
To forecast a dependent variable with regression, CB Predictor:
注意：在多元线性回归中的“多元”是指你可以有一个或多个自变量。

CB Predictor中用回归的方法预测一个因变量：

a. Creates an equation that defines the mathematical relationship between the independent variables and a dependent variable. This is the regression equation.
A、建立一个方程式来定义多个自变量和一个因变量之间的数学关系。这就是回归方程。

b. Forecasts each independent variable by running all the selected time-series forecasting methods for each one and using the best method for each.
B、通过运行所有的时间序列预测方法来预测一个因变量，并找到最合适的方法。

c. Calculates the regression equation with the forecasted independent variable values to create the forecast for the dependent variable.
C、用已预测的自变量的数据值来计算回归方程，从而预测因变量的值。

This process of creating the regression equation, forecasting the independent variables, and calculating the results to forecast the dependent variable in one easy step is called HyperCasting™.
建立回归方程的过程是，用一个简单的步骤预测因变量并计算自变量的预测结果，这个步骤称为HyperCasting™.

To use multiple linear regression: 使用多元线性回归：

1.Under Step 5 on the Data Attributes tab, if one or more variables depend on other variables that you have, select the Use Multiple Linear Regression setting.
The Regression Variables dialog appears as shown on page 106.
1、在数据属性卡的第五步，如果你有一个或多一个变量是依赖于其他变量的，选择使用多元线性回归

2.Usually the dependent variable or variables already appear in the Dependent Variables list. If they do not, follow these steps:
2、一般情况，在自变量图表中会显示自变量或者变量，如果没有，按照以下步骤操作：

a. Select the name of your dependent variable in the All Series list.
You can have more than one dependent variable. CB Predictor forecasts them all, one at a time, as functions of all the same independent variables.
A、在所有序列中选择自变量的名称
你可以有多个自变量，CB Predictor可以对他们都进行预测，如果是有相同作用的自变量，一次预测一个。

b. Click >> next to the Dependent Variables list.
The variable moves to the Dependent Variables list.
B、点击下一个自变量序列
变量会移动到下一个自变量序列

3.Verify that all independent variables are included in the Independent Variables list. If not, add them the same way:
3、确认所有的因变量都包含在因变量序列中，如果没有，用同样的方法增加：

a. Select the names of your independent variables in the All Series list.
To select multiple names, hold down either the <Ctrl> key or the <Shift> key or drag the mouse over the list.
A、在所有序列中选择因变量的名称

b. Click >> next to the Independent Variables list.
The variables move to the Independent Variables list.
B、点击下一个因变量
变量会移动到下一个自变量序列

4.To lag independent variable data by a number of time periods:
4、对因变量确定一个时间周期的数字

a. Select a variable from the Independent Variable list.
A、从因变量列表中选择一个变量

b. Enter a number in the Lag field at the bottom of the list.
B、在列表的下部周期范围空格中输入一个数字

c. Repeat for any other independent variables you want to lag.
C、对你所希望的周期重复对其他因变量进行操作

5.For any independent variables you don’t want to forecast:
5、对于不想预测的因变量：

a. Select the variable from the Independent Variable list.
A、从因变量列表中选择一个变量

b. Select the Do Not Forecast setting at the bottom of the list.
B、在序列的最下面选择不预测

c. Repeat for any other independent variables you don’t want CB Predictor to forecast.
C、对你不想进行预测的其他因变量重复进行操作

6.Click OK.
The Data Attributes tab reappears.
6、点击ok 数据对话框显示出来

7.Select the regression method to use, either standard, forward stepwise, or iterative stepwise.
7、选择回归方法，标准回归、逐步回归或者迭代回归

8.If you selected a stepwise regression, you can set settings associated with stepwise regression.
8、如果选择逐步回归，设置逐步回归的方法：

a. Click Stepwise Options.
The Stepwise Options dialog appears as shown on page 108.
A、点击逐步回归选项
逐步回归对话框显示出来

b. Set the settings. For more information on these settings, see “Regression methods”
B、进行设置，更多的设置信息查看“回归方法”

c. Click OK.
You return to the wizard.
C、点击ok 你可以返回到向导卡中

9.If you want CB Predictor to calculate the regression equation without a constant (to force the resulting equation to pass through the mathematical origin), be sure the Include Constant setting is not selected.
9、如果想用CB Predictor计算回归方程不是规定的（而不是强制的通过计算来得到方程结果），请确认包含固定设置的选项没有选择。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: 漂在生活 时间: 2009-12-21 12:22
星期一的更新，就是多啊。

作者: lanj 时间: 2009-12-22 09:44
Selecting time-series forecasting methods选择时间预测方法

You can forecast historical data using many different time-series forecasting methods. Some methods are designed to work best for certain types of data:
•Seasonal data (increasing or decreasing in a regularly recurring pattern over time)
可以用不同时间序列的预测方法来对历史数据进行预测。有些方法对不同类型的数据有特定要求。
周期数据（增加或减少在一段时间内按一定的规律变动）

•Trend data (consistently increasing or decreasing over time)
数据趋势（在一段时间内按一定的规律增加或减少）

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图3.5.png (40.77 KB, 下载次数: 58)

作者: lanj 时间: 2009-12-22 09:45
本帖最后由 lanj 于 2009-12-22 09:46 编辑

For time-series forecasting, any of the time-series forecasting methods should work with different amounts of success. However, each method has its own “specialty,” as described in the following table
对于时间序列预测，不同的时间序列预测方法有不同的特点，具体看下表：

Table 3.1 Choosing a forecasting method选择预测方法

*No trend or seasonality没有周期趋势***	*Trend only单一趋势***	*Seasonality only周期性趋势***	*Both trend and seasonality周期性单一趋势***
Single Exponential Smoothing单边指数平滑	Holt’s Double Exponential Smoothing霍尔特指数平滑	Seasonal Additive 周期递增	Holt-Winters’ Additive Holt-Winters’递增
Single Moving Average单边移动平均	Double Moving Average双边移动平均	Seasonal Multiplicative 多元周期性	Holt-Winters’ Multiplicative多元Holt-Winters

In addition to this breakdown, there are two types of seasonal methods: additive and multiplicative. Additive seasonality has a steady pattern amplitude, and multiplicative seasonality has the pattern amplitude increasing or decreasing over time.
除此之外还有两种周期性方法：加法和乘法。加法周期有一定的振动幅度，乘法周期在一段时间内有增加或者减少的振动幅度。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图3.6.png (75.4 KB, 下载次数: 53)

作者: lanj 时间: 2009-12-22 09:52
本帖最后由 lanj 于 2009-12-22 09:53 编辑

You can use a scatter plot of your data (using either Excel or the View Data button on the Input Data tab) to decide whether you have trend or seasonal data. This can help you decide which methods will work best when forecasting the data. However, selecting all the time-series forecasting methods does not significantly slow down the calculations unless you are forecasting thousands of values at once, so you might as well try them all.
你可以对数据用散点图的方法（使用excel或者查看数据按钮）来确定数据是有一定的趋势或者周期性。这也可以帮助我们在预测数据的时候选择最好的方法。否则，选择所有时间序列的预测方法等到结果可能不能确切的说明问题，除非你一次性对上千个变量进行预测，并且每个都进行尝试。

To select the forecasting methods to use: 选择预测方法：

1.       In the Data Attributes tab, click Next.
   1、在数据属性表中，点击下一步
The Method Gallery tab appears as shown in Figure 3.7. 如图3.7，方法类别会显示出来

For a description of the Method Gallery tab settings, see “Method Gallery tab”
对方法类别图表的选择，可以查看“方法类别选项”

2.In the Method Gallery tab, either:
•Try all the methods by clicking on Select All.
2、在方法集合表中
尝试所有的方法点击选择所有（select all）

•Try only some methods by clicking on Clear All and then clicking in the checkbox for each method you want to try.
只使用一种方法则清除所有，在图表中选择想要使用的方法

3.To manually set the parameters for any method, overriding the automatic calculation of parameters:
3、对各种方法手工设置参数，或者计算所有参数。

a. Double-click in the method area.
The method’s parameter dialog appears, similar to Figure A.8 on page 111.CB Predictor User Manual 69
  A、在方法列表中双击
显示出方法参数的对话框，

b. Select the User Defined setting.
The parameter fields become active. You can reset the method to automatically optimize the parameters at any time by selecting the Automatic setting.
   B、选择使用定义设置
参数变得有效率，你可以对参数进行设置，使其在任何时候选择自动优化来优化参数设置。

c. Enter the parameter values in the parameter fields.
For more information on these parameters, see “Seasonal smoothing parameters” on page 41.
      C、在参数表格中输入参数

d. Click OK.
The parameters dialog closes.
   D、点击ok
CB Predictor Note: The user-defined settings remain until you reset them. A double asterisk next to the method in the Method Gallery indicates that the method is set to use user-defined parameters.
CB Predictor注意：保持用户自定义的设置，直到您重置它们。两个星号旁边的方法表明，该方法可以自定义设置参数。

4.Set advanced settings.
For more information on setting advanced settings, see “Selecting error measures” below and “Selecting forecasting techniques” on page 70.
4、进行高级设置
高级设置的更多的信息，查看“选择错误度量”和“选择预测技术”

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

图3.7.png (109.63 KB, 下载次数: 51)

作者: lanj 时间: 2009-12-22 09:56
本帖最后由 lanj 于 2009-12-22 09:58 编辑

Selecting error measures选择误差度量

CB Predictor uses one of three error measures to determine which time-series forecasting method works best:
CB Predictor使用一个或三个偏差度量来决定那个时间序列方法更有效。

RMSE       Root Mean Squared Error 平均方根偏差
MAD       Mean Absolute Deviation 平均绝对偏差
MAPE       Mean Absolute Percentage Error 平均绝对百分比偏差

When determining the best method, CB Predictor calculates the selected error measure when fitting each method to the historical data. The method with the lowest error measure is considered best, and the rest of the methods are ranked accordingly. For more information on these error measures, see “Time-series forecasting error measures” on page 42. By default, CB Predictor uses RMSE to select the best method.
当决定了最佳的方法以后，CB Predictor会对历史数据用最适合的方法来计算误差度量。这个方法将得到最佳的最低误差，并会对其他方法进行排序。一般情况，CB Predictor选择RMSE来选择最佳的方法。

To change which error measure CB Predictor uses: 用CB Predictor对误差进行度量：

1.In the Method Gallery, click Advanced.
The Advanced Options dialog appears, similar to Figure A.9 on page 113.
1、       在方法列表中，选择高级
高级选项对话框会显示出来。

2.       Select the error measure you want CB Predictor to use to determine the best method. For more information on the Advanced Options dialog, see “Advanced Options dialog” on page 113.
2、选择误差度量，用CB Predictor来决定最佳的方法。

3.Click OK.
The Method Gallery reappears.
3、点击ok

Selecting forecasting techniques选择预测技术

CB Predictor uses one of four forecasting techniques for time-series forecasting as shown in the following table:
CB Predictor使用4种时间序列中的一种预测方法，如下表：
Table 3.2 Choosing a forecasting technique选择一种计算误差的方法

Technique

技术

Optimizes the forecasting parameters to minimize the error measure using the technique:

预测参数优化，以减少误差

Standard Forecasting

标准预测

Error measure between the fit values and the historical data for the same period.

对同一时期的隶属数据和移动值计算误差

Simple Lead

简单预测

Error measure between the historical data and the fit offset by a specified number of periods (lead).

对一个特定时期的隶属数据和移动值计算误差

Weighted Lead

加权预测

Average error measure between the historical data and the fit offset by 0, 1, 2, etc. periods, up to the specified number of periods (weighted lead).

平均之间的对历史数据和0，1，2等时期计算平均值，使其相互抵消，直至计算出特定时期误差值（加权）。

Holdout

Holdout预测

Error measure between a set of excluded data and the forecasting values. CB Predictor does not use the excluded data to calculate the forecasting parameters.

在去除的数据和预测值之间计算误差值，CB Predictor不会使用排除的数据来计算预测参数。

By default, CB Predictor uses standard forecasting to select the best method.

To change which forecasting technique CB Predictor uses:

一般情况下，CB Predictor使用标准预测作为最佳的预测方法。

1.In the Method Gallery, click Advanced.

The Advanced Options dialog appears, similar to Figure A.9 on page 113.

1、在方法列表中，选择高级

2.Select the forecasting technique you want CB Predictor to use for time-series forecasting.

For more information on the Advanced Options dialog, see “Advanced Options dialog” on page 113.

2、选择时间序列预测方法

3.If you selected Simple Lead, Weighted Lead, or Holdout, enter the appropriate lead or holdout in the field by the setting.

3、如果选择简单、加权或者holdout方法，输入合适的方法

4.Click OK.

The Method Gallery reappears.

4、点击ok

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-23 10:01
Entering the number of periods to forecast输入预测的周期数

After CB Predictor finds the method that best fits the historical data, it is ready to use that same method to forecast future values. You need to decide how many time periods you want to forecast.
用CB Predictor找到最适合历史数据的统计方法后，可以用同样的方法预测未来的数据。你需要确定你想要预测多长周期的数据。

As you are trying to decide, there are a few points to keep in mind:
用CB Predictor找到最适合历史数据的预测方法后，

•The first few values are fairly reliable. Only forecast as many values as you need.
开始的值是非常可靠的，根据需要预测更多的值。

•The farther out you try to forecast, the less reliable the forecasted values are. The confidence interval of any forecast grows to reflect this decrease in reliability.
如果需要预测更多的数据，预测值就更不稳定。预测的置信区间反映减少趋势的稳定性。

•The maximum number of time periods possible is 100.
时间周期的最大值是100

To enter the number of periods to forecast: 输入预测的周期数：

1.In the Method Gallery, click Next.
The Results tab appears as shown in Figure A.10 on page 115.
1、在方法图表中，点击下一步

2.Under Step 7, enter the number of periods you want to forecast.
For details, see “Results tab” beginning on page 115.
2、在第七步，输入需要预测的周期数

Selecting a confidence interval选择置信区间

The confidence interval defines the range above and below a forecasted value where the value has some probability of occurring. For example, a confidence interval of 10% and 90% gives two points for each forecasted value. The lower point represents the 10th percentile. The higher point represents the 90th percentile. The farther out the forecast is, the larger this range is.
置信区间是定义每个预测值可能发生的上限和下限的范围。例如，如果对每个预测值给出置信区间是10%和90%两个值。下限是10%，上限是90%。

To set a confidence interval, under Step 8 on the Results tab, select the confidence interval you want to calculate and display with your results. For details, see “Results tab”
根据第八步设置置信区间，选择需要计算的置信区间，并得出结果。如果需要了解更详细的信息，查看“结果列表”。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-23 10:03
Select your results选择结果

After you identify your historical data, select the methods to try, and decide how many time periods you want to forecast, you must decide which results you want to create. CB Predictor can:
确认历史数据之后，选择预测的方法，并决定需要预测多少周期的数据，并且决定需要生成怎样的预测报告。CB Predictor可以生成以下几种报告：

•Paste the forecasted data anywhere on your worksheet or into a new worksheet
   在工作表或者新的sheet表中粘贴预测结果

•Create a chart that can show historical data, fitted values, and forecasted data and its confidence interval
生成一个历史数据的图表，图表中会显示历史数据的最佳值、预测值和置信区间

•Generate a report summarizing the findings
生成一份总结结果的报告

•Create a PivotTable of all the historical data, fitted values, forecasted data, and confidence intervals
生成包含所有历史数据的数据表，图表中包含了历史数据的最佳值、预测值和置信区间。

•Create a PivotTable of some or all the method information for each forecast, including the errors, parameters, and statistics for each method tried
生成一份对每次预测的所有或者部分的预测方法的信息，包括每种预测方法的误差、参数和统计方法。

You can choose to generate only basic results, such as pasted data and charts, until you are sure you have the forecasts you want. Then you can create customized, detailed reports and tables to present to others.
可以选择只生成基础的结果，比如粘贴数据和图表，以及其他需要的预测结果。也可以定制有特定结果和图表的报告。

To generate the results you want: 获得你需要的结果：

1.Under Step 9 on the Results tab (page 115), to paste forecasted values:
a. Select the Paste Forecasts setting.
Enter the cell reference of the starting location in the field to the right. The default starting location pastes the forecasted values immediately at the end of your historical data.
b. To paste the data somewhere other than the end of the data, enter the cell where you want the pasted data to start or click Select to select a different cell interactively.
c. Select to paste the forecasted values in rows or columns.
1、按照第九步，粘贴预测结果
A、选择粘贴结果设置
在图表的右边输入粘贴结果的单元格位置，默认情况是在历史数据的后面粘贴预测结果。
B、在数据的结尾粘贴预测结果，或者选择要粘贴的单元格的范围
C、选择预测结果以行或者列的形式粘贴

2.Select the other types of output you want by clicking on the appropriate checkboxes.
2、在列表中选择想要输出的结果类型

3.To customize the output types:
a. Click Preferences.
The Preferences dialog appears as shown on page 117.
b. Click the tab of the result you want to customize.
c. Change the settings.
For more information on these settings, see “Preferences dialog” on page 117.
d. Repeat steps b and c for any other result you want to customize.
e. Click OK.
The Results tab of the wizard reappears.
3、定制输出的类型：
  A、选择偏好
  B、点击需要定制的结果
  C、修改设置
  D、对其他预测也重复操作
  E、点击ok

4.Enter a title in the Title field to identify your output.
This title appears on all the result sheets that CB Predictor creates. The default title is the worksheet name.
4、输入预测结果的标题
这个标题会显示在所有的预测结果上，一般默认的标题是sheet表的名称。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

作者: lanj 时间: 2009-12-23 10:04
Previewing and running the forecast预览并运行结果

CB Predictor actually calculates the forecasts for each data series when you select either Preview or Run. The difference is that Preview lets you view the forecasts and make changes before it generates any of the results. Run simply generates the results.
当选择运行预测结果时，CB Predictor会对每组数据的预测结果进行计算，不同的是预览会在得出结果之前让你查看预测的情况并作出改变，而运行只是简单的得到结果而已。

1.To preview the forecast before generating the selected results, click Preview.
The Preview dialog appears. For more information, see “Preview Forecast dialog”
1、在结果生成之前要进行预览，点击预览

2.To run the forecast and produce the selected results, click Run.
2、要得到最后的预测结果，点击运行

CB Predictor Note: You can click Preview or Run from any of the wizard tabs at any time, as long as the data have been properly defined in the Input Data tab.
CB Predictor注意：在数据输入端过程中，你可以随时运行和预览就结果。

-----------------------------------------------------------------------
译者权归“思步网（www.step365.com）”及“lanj（西北偏北）”共同所有，未经许可，请勿转载

欢迎光临思步网 (http://www.step365.com/)