MLS-C01 Exploratory Data Analysis Practice Question

Correct answer & explanation

The most appropriate technique for identifying the most important features for predicting a binary target is to train a Random Forest classifier and use the built-in feature_importances_ attribute (Option C). Random Forest is a supervised ensemble method that provides a ranking of feature importance based on how much each feature reduces impurity (e.g., Gini impurity) across all trees. Option A (t-SNE) is a nonlinear dimensionality reduction technique primarily used for visualization in 2D/3D; it does not provide feature importance. Option B (K-means clustering) is an unsupervised clustering algorithm that does not use the target variable and cannot identify predictive features. Option D (PCA) is an unsupervised dimensionality reduction method that finds principal components maximizing variance, but these components are not directly interpretable as feature importance for a specific target variable.