Machine Learning - Evaluate

Question 1

Cross-Validate Model:

Answer 1

Cross-validates parameter estimates for classification or regression models by partitioning the data.

Use the Cross-Validate Model module if you want to test the validity of your training set and the model. Cross-validation partitions the data into folds, and then tests multiple models on combinations of folds.

Question 2

Evaluate Model:

Answer 2

Evaluates a scored classification or regression model by using standard metrics.

In most cases, you’ll use the generic Evaluate Model module. This is especially true if your model is based on one of the supported classification or regression algorithms.

Question 3

Evaluate Recommender:

Answer 3

Evaluates the accuracy of recommender model predictions.

For recommendation models, use the Evaluate Recommender module.