2.6 Types of Machine Learning

ML has four major types. Each type solves different problems.

⭐ Type 1 — Supervised Learning (Most Important in ML)

Supervised learning uses labeled data.

Example:

Used for:

• Email spam detection
• Disease prediction
• Face recognition
• Credit card fraud detection

Why it’s powerful:

It learns the relationship between input → output.

⭐ Type 2 — Unsupervised Learning

Learns from unlabeled data. Finds hidden patterns.

Examples:

• Customer segmentation
• Product grouping
• Document clustering

Algorithms:

• K-Means
• Hierarchical Clustering
• PCA

⭐ Type 3 — Semi-Supervised Learning

Mix of:

• Few labeled samples
• Many unlabeled samples

Used in:

• YouTube video tagging
• Facial recognition
• Speech labeling

⭐ Type 4 — Reinforcement Learning (Learning by Reward)

Agent takes actions → gets reward → learns.

Used in:

• Self-driving cars
• Robotics
• Game AI
• Navigation