This command is available with the Predictive Analytics Module. Click here for more information about how to activate the module.
When you use Discover Key Predictors to remove the least important predictors, Minitab Statistical Software produces results for the model with the best value of the accuracy criterion for the analysis, such as the minimum average –loglikelihood. Minitab lets you explore other models from the sequence that led to the identification of the optimal model. Generally, you select an alternative model if another model has a value of the criterion close to the best, but with fewer predictors. A model with fewer predictors is easier to interpret, can have better prediction accuracy, and allows you work with a smaller number of predictors.
For example, the following model selection table has 13 steps. The model with the least average –loglikelihood is the model with all of the predictors. The model at step 11 has an average –loglikelihood that is relatively close to the best value. The model at step 11 has 3 predictors. The full results from the model at step 11 are also of interest.
Click Select an Alternative Model in the output. A dialog box opens that shows a plot of the criterion against the number of eliminated predictors and a table that summarizes the steps.
To select an alterative model, click a point on the graph or a row in the table. Press Display results to create the results for that model.
Once you display the results, you can click a button in the output to tune the hyperparameters of the model or to make predictions from the model. For more information, go to Select hyperparameter values to evaluate from the results for TreeNet® Classification or Predict new results for Fit Model and Discover Key Predictors with TreeNet® Classification.
To compare the output of two different analyses or reports, right-click the second item in the Navigator and choose Open in Split View.