Supervised Learning

The machine learning task of learning a function that maps an input to an output based on example input-output pairs.

How it works

In supervised learning, a labelled training dataset contains input-output pairs (x, y). An algorithm learns a function f such that f(x) ≈ y by minimising a loss function that measures prediction error. For classification, the label y is a category; for regression, it is a continuous value. The model is validated on a held-out set to check generalisation. Common supervised algorithms include linear regression, logistic regression, SVMs, decision trees, and neural networks.

Why it matters

Supervised learning is the most widely deployed form of machine learning because it is the most direct: you show the model exactly what you want it to predict. It powers spam filters, fraud detectors, medical diagnostic tools, search engine ranking, and virtually every industrial prediction system. Understanding supervised learning is the prerequisite for understanding almost every practical ML application, and the challenges it presents — data labelling costs, distribution shift, class imbalance — are the dominant practical constraints in production ML.