202502012201
tags: #machine-learning #supervised-learning

Regression predicts continuous numerical values, while classification predicts discrete categories or classes.

Regression Examples:

• Predicting house prices (output: dollar amount)
• Forecasting temperature (output: degrees)
• Estimating stock prices (output: price value)

Classification Examples:

• Email spam detection (output: spam/not spam)
• Image recognition (output: cat/dog/bird)
• Medical diagnosis (output: disease/healthy)

Both types use Cost Function to measure performance, but regression typically uses mean squared error while classification uses log-likelihood or cross-entropy loss.

The choice affects which algorithms and Optimization techniques work best for your problem.

Reference

Machine Learning Yearning by Andrew Ng