Lasso regression also adds a penalty term (L1 regularization) to the linear regression cost function. It not only prevents overfitting but also tends to force some coefficients to be exactly zero, effectively performing feature selection.

👉 What it does: Adds penalty that can shrink some coefficients to exactly zero → automatic feature selection

✅ When to use:

• You want a simpler model
• You suspect only a few features actually matter
• You need interpretability + feature selection

🧠 Real-life examples:

• Identifying key factors affecting house prices (maybe only 5 out of 50 matter)
• Marketing: finding which campaigns actually drive sales
• Genomics: selecting important genes from thousands

💡 Key idea: 👉 Eliminates irrelevant features