📈 01 — Regression Overview

Regression predicts a continuous numeric value.

Examples:

• house price
• sales revenue
• delivery time
• customer lifetime value

---

Basic Workflow

X = df.drop(columns=["price"])
y = df["price"]

Then:

1. split data
2. preprocess features
3. train regression model
4. evaluate error
5. inspect residuals

---

Common Regression Models

Model	Strength
Linear Regression	simple baseline, interpretable
Ridge	handles multicollinearity
Lasso	can shrink features to zero
Decision Tree Regressor	captures nonlinear patterns
Random Forest Regressor	strong general-purpose model
Gradient Boosting	often high accuracy

---

Key Idea

Regression is not about being exactly correct. It is about minimizing prediction error and understanding when errors are acceptable.

---

Next

➡️ 02-linear-regression