Knowledge Graph — Coursera Notes › Academic disciplines › Computer Science / Information Technology › Artificial Intelligence › Machine Learning & Data

Feature selection

concept · part of Machine Learning & Data

Technique to improve model performance by retaining relevant features and discarding irrelevant ones.

Inside Feature selection (3)

• Backward elimination — Starts with all features and iteratively removes the least significant one based on p-values.
• Forward selection — Starts with no features and iteratively adds the most significant one based on correlation or model improvement.
• LASSO — Regularization technique that performs automatic feature selection by shrinking coefficients to zero via L1 penalty.

Connections

• Uses Backward elimination
• Uses Forward selection
• Uses LASSO

This is the text view of an interactive 3D knowledge graph — open this page with JavaScript enabled to explore it visually.