Stacking and blending in ML Python - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the stacking classifier from scikit-learn.

ML Python

from sklearn.ensemble import [1]

Drag options to blanks, or click blank then click option'

AGradientBoostingClassifier

BStackingClassifier

CRandomForestClassifier

DVotingClassifier

Attempts:

3 left

💡 Hint

Common Mistakes

Importing VotingClassifier instead of StackingClassifier.

Importing RandomForestClassifier which is a single model, not stacking.

The StackingClassifier is the correct class to import for stacking models in scikit-learn.

2fill in blank

medium

Complete the code to create a stacking classifier with logistic regression as the final estimator.

ML Python

stack = StackingClassifier(estimators=estimators, final_estimator=[1]())

Drag options to blanks, or click blank then click option'

ADecisionTreeClassifier

BRandomForestClassifier

CKNeighborsClassifier

DLogisticRegression

Attempts:

3 left

💡 Hint

Common Mistakes

Using a complex model like RandomForestClassifier as final estimator.

Using a base estimator instead of a final estimator.

LogisticRegression is commonly used as the final estimator in stacking to combine base model predictions.

3fill in blank

hard

Fix the error in the code to correctly fit the stacking classifier on training data.

ML Python

stacking_model = StackingClassifier(estimators=estimators, final_estimator=LogisticRegression())
stacking_model.[1](X_train, y_train)

Drag options to blanks, or click blank then click option'

Afit

Bpredict

Ctransform

Dscore

Attempts:

3 left

💡 Hint

Common Mistakes

Calling predict before fitting the model.

Using transform which is for feature processing.

The fit method trains the stacking model on the training data.

4fill in blank

hard

Fill both blanks to create a blending dataset by training base models and collecting their predictions.

ML Python

blend_data = np.zeros((X_val.shape[0], len(estimators)))
for i, (name, model) in enumerate(estimators):
    model.[1](X_train, y_train)
    blend_data[:, i] = model.[2](X_val)

Drag options to blanks, or click blank then click option'

Afit

Bpredict

Ctransform

Dscore

Attempts:

3 left

💡 Hint

Common Mistakes

Using transform instead of predict for collecting predictions.

Calling score instead of fit to train models.

First, each base model is trained with fit, then predictions are collected with predict to form the blending dataset.

5fill in blank

hard

Fill all three blanks to define a stacking classifier with two base models and a final estimator.

ML Python

estimators = [("rf", RandomForestClassifier()), ("svc", [1]())]
stack = StackingClassifier(estimators=estimators, final_estimator=[2]())
stack.fit([3], y_train)

Drag options to blanks, or click blank then click option'

ASVC

BLogisticRegression

CX_train

DDecisionTreeClassifier

Attempts:

3 left

💡 Hint

Common Mistakes

Using DecisionTreeClassifier instead of SVC as base model.

Passing y_train instead of X_train to fit.

SVC is used as a base model, LogisticRegression as the final estimator, and X_train is the training data to fit the model.