In Regression Learner, use the response plot to try to identify predictors that are useful for predicting the response. To visualize the relation between different predictors and the response, under X-axis, select different variables in the X list.
Before you train a regression model, the response plot shows the training data. If you have trained a regression model, then the response plot also shows the model predictions.
Observe which variables are associated most clearly with the response. When you
carbig data set, the predictor
Horsepower shows a clear negative association with the
Look for features that do not seem to have any association with the response and use Feature Selection to remove those features from the set of used predictors.
You can export the response plots you create in the app to figures. See Export Plots in Regression Learner App.
In Regression Learner, you can specify different features (or predictors) to include in the model. See if you can improve models by removing features with low predictive power. If data collection is expensive or difficult, you might prefer a model that performs satisfactorily with fewer predictors.
On the Regression Learner tab, in the Features section, click Feature Selection .
In the Feature Selection dialog box, clear the check boxes for the predictors you want to exclude, and then click OK.
Click Train to train a new model using the new predictor options.
Observe the new model in the Models pane. The Current Model Summary pane displays how many predictors are excluded.
Check which predictors are included in a trained model. Click the model in the Models pane and look at the check boxes in the Feature Selection window.
Try to improve the model by including different features.
For an example using feature selection, see Train Regression Trees Using Regression Learner App.
Use principal component analysis (PCA) to reduce the dimensionality of the predictor space. Reducing the dimensionality can create regression models in Regression Learner that help prevent overfitting. PCA linearly transforms predictors to remove redundant dimensions, and generates a new set of variables called principal components.
On the Regression Learner tab, in the Features section, select PCA.
In the Advanced PCA Options dialog box, select the Enable PCA check box, and then click OK.
Click Train again. The
function transforms your selected features before training the
By default, PCA keeps only the components that explain 95% of the variance. In the Advanced PCA Options dialog box, you can change the percentage of variance to explain by selecting the Explained variance value. A higher value risks overfitting, while a lower value risks removing useful dimensions.
Manually limit the number of PCA components. In the
Component reduction criterion list, select
Specify number of components. Select the
Number of numeric components value. The number
of components cannot be larger than the number of numeric predictors.
PCA is not applied to categorical predictors.
You can check PCA Options for trained models in the Current Model Summary pane. For example:
PCA is keeping enough components to explain 95% variance. After training, 2 components were kept. Explained variance per component (in order): 92.5%, 5.3%, 1.7%, 0.5%
To learn more about how Regression Learner applies PCA to your data, generate code
for your trained regression model. For more information on PCA, see the