Practical 4: Classification with Logistic Regression

Objective

Build binary and multi-class classification models using logistic regression.

Duration

3-4 hours

Prerequisites

• Practicals 1-3 completed
• Understanding of classification

What You’ll Learn

• ✅ Implement logistic regression
• ✅ Handle binary and multi-class classification
• ✅ Create confusion matrices
• ✅ Calculate precision, recall, F1-score
• ✅ Plot ROC curves

📋 Tasks

1. Build Classification Model

from sklearn.linear_model import LogisticRegression

model = LogisticRegression(max_iter=1000)
model.fit(X_train, y_train)
y_pred = model.predict(X_test)

2. Evaluate Classification

from sklearn.metrics import confusion_matrix, classification_report, roc_auc_score

cm = confusion_matrix(y_test, y_pred)
print(classification_report(y_test, y_pred))

auc = roc_auc_score(y_test, model.predict_proba(X_test)[:, 1])
print(f"AUC Score: {auc:.4f}")

📊 Learning Outcomes

• Build logistic regression classifier
• Generate confusion matrix
• Calculate classification metrics
• Plot ROC curves
• Interpret results