Feature importance explanation in ML Python - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the function that calculates feature importance from a trained tree model.

ML Python

from sklearn.ensemble import RandomForestClassifier
model = RandomForestClassifier()
model.fit(X_train, y_train)
importance = model.[1]

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

Afit

Bpredict

Cscore

Dfeature_importances_

Attempts:

3 left

💡 Hint

Common Mistakes

Using the method 'predict' instead of the attribute for importance.

Trying to call 'fit' again instead of accessing importance.

Using 'score' which returns accuracy, not feature importance.

The feature_importances_ attribute of a trained RandomForestClassifier gives the importance scores of each feature.

2fill in blank

medium

Complete the code to plot the feature importances using matplotlib.

ML Python

import matplotlib.pyplot as plt
features = ['age', 'income', 'score']
importances = model.feature_importances_
plt.bar(features, [1])
plt.show()

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

Aimportances

Brange(len(features))

Cmodel

Dfeatures

Attempts:

3 left

💡 Hint

Common Mistakes

Using feature names as bar heights instead of importance values.

Passing the model object instead of importance values.

Using range indices instead of actual importance values.

The bar heights should be the feature importance values stored in importances.

3fill in blank

hard

Fix the error in the code to get feature importance from a trained linear model.

ML Python

from sklearn.linear_model import LogisticRegression
model = LogisticRegression()
model.fit(X_train, y_train)
importance = model.[1]

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

Aintercept_

Bcoef_

Cpredict_proba

Dfeature_importances_

Attempts:

3 left

💡 Hint

Common Mistakes

Trying to use feature_importances_ which is for tree models only.

Using predict_proba which is a method, not importance.

Using intercept_ which is the bias term, not feature importance.

Linear models like LogisticRegression use coef_ to represent feature importance as coefficients.

4fill in blank

hard

Fill both blanks to create a dictionary mapping features to their importance scores, filtering only those with importance greater than 0.1.

ML Python

important_features = {feature: importance for feature, importance in zip([1], [2]) if importance > 0.1}

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

Afeatures

Bimportances

Cmodel

DX_train.columns

Attempts:

3 left

💡 Hint

Common Mistakes

Using the model object instead of feature names or importance values.

Swapping the order of features and importances in zip.

Using dataset columns without matching importance values.

We zip feature names with their importance values and filter by importance > 0.1.

5fill in blank

hard

Fill all three blanks to compute and print the top 3 features by importance from a trained RandomForest model.

ML Python

import numpy as np
indices = np.argsort(model.[1])[::-1]
top_features = [features[i] for i in indices[:[2]]]
top_importances = [model.feature_importances_[i] for i in indices[:[3]]]
for f, imp in zip(top_features, top_importances):
    print(f"Feature: {f}, Importance: {imp:.2f}")

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

Afeature_importances_

B3

Dcoef_

Attempts:

3 left

💡 Hint

Common Mistakes

Using coef_ which is not available for RandomForest.

Using different numbers for second and third blanks causing mismatch.

Not reversing the sorted indices to get descending order.

We sort the feature_importances_ descending, then select top 3 features and their importances to print.