Machine Learning

Supervised Learning

ML, AI, Data Science, Deep Learning — What’s What?

• Data Science = everything data: collection, cleaning, analysis, visualization.
• Soft: dashboards, simple stats, reporting.
• Hard: ML modeling, algorithmic analysis.

• AI = broader goal of creating intelligent systems.
• May or may not use learning.
• Think planning, robotics, decision-making.

• Machine Learning (ML) ⊂ Data Science ⊂ AI
Uses data to learn patterns and make predictions.

• Deep Learning (DL) ⊂ ML
Neural networks for complex data (images, text, audio).

Learning Types

• Supervised: Learn from labeled data
→ classification (A/B/C) or regression (real numbers)

• Unsupervised: No labels
→ clustering, anomaly detection, dimensionality reduction

• Reinforcement: Learn from rewards/punishment
→ used in robotics, games, adaptive control

Model Types

• Parametric: fixed structure (e.g., linear regression)

• Non-parametric: flexible, data-driven (e.g., k-NN)