Evaluate performance

Lastly, and as always, we want to evaluate performance of our best model to check how well or poorly we are doing. Ideally it's best to do back-testing, but that's an involved process we don't have room to cover in this course.

We've already seen the R\(^2\) scores, but let's take a look at the scatter plot of predictions vs actual results using matplotlib. Perfect predictions would be a diagonal line from the lower left to the upper right.

Use the best number for max_features in our RandomForestRegressor (rfr) that we found in the previous exercise (it was 4).
Make predictions using the model with the train_features and test_features.
Scatter actual targets (train/test_targets) vs the predictions (train/test_predictions), and label the datasets train and test.

Exercise

Evaluate performance

Instructions