IMS Control Region - Control Parameters
Figure 9-12 shows the control parameters used to generate the final multilinear regression model of this case study.
rm
Figure 9-12 shows the control parameters used to generate the final multilinear regression model of this case study.
Figure 9-12. IMS Control Region Case Study Control Screen
/-------------------- Multivariate Regression Forecasting --------------------\ |Command ===> | |Enter a ? in any data entry field for more information on valid values. | |Composing CA MICS Inquiry: IMSOVH - IMS Control Region Overhead Study | | | |Report title ===> Capture ratios on test workload | | | |Selection criteria: | | Dependent file ===> IMS - IMS Analysis File | | Dependent element ===> CTLCPUTM | | Start date ===> _________ (ddmonyyyy) | | SYSID ===> SYS2 | | Zone ===> 1_______ | | No. weeks ===> ____ (1-9999) | | Specify CAPAPU Values ===> N (Y/N) | |Save forecast ===> ___ (YES/NO/AGE) | |Independent file ===> IMS - IMS Analysis File | |Independent elements ===> MPPCPUTM BMPCPUTM ________ ________ ________ | | ________ ________ ________ ________ ________ | | | |Confidence limits (percent) ===> __ (70/90/95) | |Min R-square improvement ===> 0.001 (0.000-0.100) | |Delete observations ===> N (Y/N/R) | | | --------------------------------------------------------------------------------
Note that the same resource file (IMS) was used for both the dependent variable (CTLCPUTM) and the two independent variables (MPPCPUTM and BMPCPUTM). Note also that the analyst previously generated and saved forecasts for each of the independent variables, using one of the forecasting routines of this component. This is required for Multivariate Regression Forecasting to produce forecasts of the dependent variable.
In this case study, a Minimum R-squared improvement value of .001 is used for the final model. Sometimes in the evaluation of multilinear models you may feel that the default value of .05 does not quite provide a fine enough distinction between models to produce the best multilinear regression model. Normally, a value of .01 should be sufficient to make this distinction.
In this study, the first execution of the program uses the default value for Minimum R-squared improvement. However, this value is not fine enough to distinguish between the two models listed in the Model Analysis Report, shown in Figure 9-13 in IMS Control Region - Model Analysis Report. When this type of situation occurs, the algorithm chooses the first model, because the second model failed to provide the Minimum R-squared improvement you specified. The model that the stepwise regression process chooses is therefore the one which uses only MPPCPUTM as the independent variable. In the example, the analyst decided to see the results of a model that chose both MPPCPUTM and BMPCPUTM as independent variables before committing to one or the other. Therefore the program was rerun with a lower Minimum R-squared improvement value to force the choice of the second model.