Building the second-layer classifier

Now you'll work on the next two steps of the process.

Step 3: Append the predictions to the dataset The predictions with the first-layer estimators were already calculated and are available to you as: pred_dt and pred_knn. You'll use these to create a DataFrame and append them to the training features.

Step 4: Build the second-layer meta estimator For this purpose you'll build the default DecisionTreeClassifier. This will take as input feature the concatenation of the original training features and the predictions from the first-layer estimators.

Create a pandas DataFrame with the predictions from the first-layer estimators: pred_dt and pred_knn (specifying the appropriate keys and order).
Concatenate the training features set with the predictions DataFrame using axis = 1.
Build the second-layer meta estimator: a DecisionTreeClassifier with random_state = 500, and fit it using the concatenated features set.

Exercise

Building the second-layer classifier

Instructions