The Model Analysis Report corresponding to this case study is presented in Figure 10-14. Section 10.3.2.1 contains a detailed discussion of each of the fields appearing on this report. Several points should be noted in this case study. At first glance the values of r-squared, F, and p all suggest that this is a solid model of the relationship between EXCPs and the specified business elements. In fact, the r-squared value indicates that the proposed model explains 98% of the variability in the historical data. However, there are negative coefficients associated with the business elements included in the model. If the model is truly valid, these would suggest that, as the business element volumes increase, the expected resource consumption of the application decreases. This does not make sense and you should reject the proposed model for this reason. You should investigate the historical data to determine if reruns or other factors caused anomalies in the historical data.
CA MICS Capacity Planner ANALYSIS OF BUSINESS ELEMENT MODELS MODEL OF: BILEXCPS BASED ON: INVOICES UPDATES OPN_ITEM ----------BUSINESS-ELEMENTS---------- R**2 F P INTERCEPT NAME COEFFICIENT F P -------- ------- ----- --------- -------- ----------- ------- ----- 0.78 17.26 .0089 7768.21 INVOICES -0.159782 17.26 .0089 0.91 20.86 .0077 9091.08 INVOICES -0.14591 28.35 .0060 OPN_ITEM -.0328758 6.27 .0665 0.98 44.36 .0055 7938.98 INVOICES -0.122026 47.07 .0063 UPDATES 0.196209 8.91 .0584 OPN_ITEM -.0309463 16.42 .0271
Figure 10-14. Model Analysis Report The Residual Analysis Report details how the business element-based model proposed by the analyst fits the historical data series. The Residual Analysis Report corresponding to the case study is presented in Figure 10-15 Section 10.3.2.2 contains a detailed discussion of each of the fields appearing on the standard report. Generally, a good model has no pattern in the plot of the residual values. The presence of a pattern in the residuals often suggests that there is some systematic pattern to the historical data that is not being accounted for by the model. For this example, the residuals are very small (the residuals are plotted on a scale from -58 to +58, while the observed values range from 3595 to 4268), so the errors in the predicted values are on the order of 1%. Also, there seems to be no systematic pattern to the residuals. These observations support the conclusion drawn from the model analysis report that, on a purely statistical basis, the model appears valid. Confidence limits are reported on the right side of the Residual Analysis Report. For a business element model, these limits are a function of the estimates for future business elements rather than simply increasing monotonically as was the case with models developed by Univariate Model Forecasting. Note that the confidence limits for the August forecast are small since the business element estimates provided to the model were in the minimum to maximum range observed in the historical data. However, the confidence limits for September and October are larger since the business element estimates are outside of the range of observed historical values.
C A M I C S C A P A C I T Y P L A N N E R BUSINESS ELEMENT FORECAST AND RESIDUAL ANALYSIS BUSINESS ELEMENT BASED FORECAST OF: BILEXCPS MODEL BASED ON BUSINESS ELEMENTS : INVOICES UPDATES OPN_ITEM -BUSINESS-ELEMENTS- PREDICTED RESIDUAL -------------PLOT OF RESIDUALS------------- CONFIDENCE DATE BILEXCPS NAME VALUE BILEXCPS BILEXCPS -58 0.0 +58 LIMITS +/- ------- ---------- -------- --------- ---------- ---------- ------------------------------------------- ---------- 31JAN98 4205.00 INVOICES 24017 4147.62 57.38 | 0+++++++++++++++++++ | 172.66 UPDATES 3145 | 0 | OPN_ITEM 47752 | 0 | | 0 | 28FEB98 4268.00 INVOICES 21570 4292.32 -24.32 | --------0 | 145.01 UPDATES 2960 | 0 | OPN_ITEM 51552 | 0 | | 0 | 31MAR98 3970.00 INVOICES 23411 3932.92 37.08 | 0++++++++++++ | 133.57 UPDATES 2012 | 0 | OPN_ITEM 49896 | 0 | | 0 | 30APR98 4384.00 INVOICES 21644 4391.34 -7.34 | --0 | 171.87 UPDATES 2500 | 0 | OPN_ITEM 45144 | 0 | | 0 | 31MAY98 3908.00 INVOICES 23476 3961.48 -53.48 | ------------------0 | 115.22 UPDATES 2709 | 0 | OPN_ITEM 53136 | 0 | | 0 | 30JUN98 3595.00 INVOICES 26311 3644.67 -49.67 | -----------------0 | 171.37 UPDATES 2108 | 0 | OPN_ITEM 48384 | 0 | | 0 | 31JUL98 3651.00 INVOICES 24774 3610.66 40.34 | 0+++++++++++++ | 157.48 UPDATES 2103 | 0 | OPN_ITEM 55512 | 0 | | 0 | 31AUG98 . INVOICES 25000 3764.39 . | 0 | 133.61 UPDATES 2000 | 0 | OPN_ITEM 49000 | 0 | | 0 | 30SEP98 . INVOICES 27000 3483.33 . | 0 | 224.90 UPDATES 2600 | 0 | OPN_ITEM 54000 | 0 | | 0 | 31OCT98 . INVOICES 23000 3893.67 . | 0 | 292.69 UPDATES 3150 | 0 | OPN_ITEM 60000 | 0 |
Figure 10-15. Residual Analysis Report
|
Copyright © 2014 CA.
All rights reserved.
|
|