Feature Engineering for Modern Machine Learning with Scikit-Learn

6.3 Practical Exercises: Chapter 6

In this exercise section, we’ll apply regularization techniques for feature selection using Lasso and Ridge. These exercises will help solidify your understanding of L1 and L2 regularization and hyperparameter tuning.

Exercise 1: Applying Lasso for Feature Selection

Objective: Use Lasso regression to identify the most important features from a dataset and observe how changing the alpha parameter affects feature selection.

Instructions:

1. Load a dataset with at least 15 features.
2. Apply Lasso regression and experiment with different values of alpha.
3. List the non-zero coefficients (selected features) for each alpha value and plot them to visualize which features remain relevant as alpha increases.

Solution:

from sklearn ...