Machine Learning
Teaching computers to learn from examples, not rules.
Machine Learning (ML) is a branch of AI where systems improve by studying patterns in data — without being explicitly programmed for each scenario.
Instead of writing rules like "if price drops 10% in 5 days, sell," you feed historical data with labels ("went up" / "went down") and the model discovers its own patterns.
Three core flavours: Supervised learning (learns from labeled examples), Unsupervised learning (finds hidden clusters), and Reinforcement learning (learns via rewards and penalties). Most tools you use today — spam filters, recommendation engines, image recognition — run on supervised ML.
In plain terms
Like training a dog: you reward correct behaviour repeatedly until the dog generalises the rule on its own — no instruction manual required.