🧭 01 — Classification Overview

Classification predicts a category.

Examples:

• spam vs not spam
• churn vs not churn
• fraud vs legitimate
• disease vs no disease
• sentiment: positive, neutral, negative

---

Binary vs Multiclass

Type	Example
Binary	churn yes/no
Multiclass	low/medium/high risk
Multilabel	movie can be comedy and drama

---

Classification Workflow

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

1. split with stratification
2. preprocess features
3. train classifier
4. predict labels/probabilities
5. evaluate with suitable metrics
6. inspect confusion matrix and errors

---

Common Models

Model	Good For
Logistic Regression	interpretable baseline
KNN	simple distance-based classification
Naive Bayes	text and simple probabilistic tasks
Decision Tree	interpretable nonlinear rules
Random Forest	strong general-purpose baseline
Gradient Boosting	high-performing tabular model

---

Next

➡️ 02-logistic-regression