1. Learn
    2. /
  2. Courses
    3. /
  3. Advanced Deep Learning with Keras

Exercise

Evaluate the model on a test set

After fitting the model, you can evaluate it on new data. You will give the model a new X matrix (also called test data), allow it to make predictions, and then compare to the known y variable (also called target data).

In this case, you'll use data from the post-season tournament to evaluate your model. The tournament games happen after the regular season games you used to train our model, and are therefore a good evaluation of how well your model performs out-of-sample.

The games_tourney_test DataFrame along with the fitted model object is available in your workspace.

Instructions

100 XP
  • Assign the test data (seed_diff column) to X_test.
  • Assign the target data (score_diff column) to y_test.
  • Evaluate the model on X_test and y_test.