scikit-learn Cookbook , Second Edition - Second Edition

How to do it...

1. Begin by scaling the input variables. Train the scaler only on the training data:

from sklearn.preprocessing import StandardScalerscaler = StandardScaler()scaler.fit(X_train)X_train_scaled = scaler.transform(X_train)

2. Then, perform the scaling on the test set:

X_test_scaled = scaler.transform(X_test)

3. Finally, perform a randomized search (or grid search if you prefer) to find a reasonable value for alpha, one that scores well:

from sklearn.model_selection import RandomizedSearchCVfrom sklearn.neural_network import MLPRegressorparam_grid = {'alpha': [10,1,0.1,0.01],               'hidden_layer_sizes' : [(50,50,50),(50,50,50,50,50)],               'activation': ['relu','logistic'],               'solver' : ['adam']               }pre_gs_inst = RandomizedSearchCV(MLPRegressor(random_state=7), ...