Skip to main content

Machine Learning — Fundamentals (Level 3)

A conceptual introduction to Machine Learning (ML) core ideas.

What is Machine Learning?

Machine Learning is a subfield of AI where systems learn patterns from data rather than being explicitly programmed for each case. The goal is to make predictions or discover structure.

Common problem types

  • Supervised learning: learn from labelled examples (e.g., spam vs not-spam). Output: labels or values.
  • Unsupervised learning: find structure in unlabelled data (e.g., clustering customers).
  • Reinforcement learning: an agent learns by interacting with an environment and receiving rewards.

Typical pipeline (high-level)

  1. Collect data
  2. Clean and pre-process data
  3. Choose a model/algorithm
  4. Train the model on data
  5. Evaluate performance on held-out data
  6. Deploy and monitor

Simple example (intuition)

Predicting house price: model looks for links between features (size, location) and price from past sales, then uses that to estimate new house prices.

Datasets and features

  • Features are the inputs (age, pixels, temperature).
  • Labels are the outputs we want to predict (class, price).

Short practical exercise

  1. Use an online ML demo (scikit-learn or Google Teachable Machine) and try training a small classifier. Note accuracy on training vs test set and explain why they differ.