Regression is a machine learning task where a model predicts a continuous numerical value, rather than choosing a category. It answers “how much?” or “how many?”
What it means in plain English
Where classification predicts a category, regression predicts a number on a continuous scale — a price, a temperature, a score, an amount. The model learns the relationship between input features and the numerical outcome from labelled examples, then predicts values for new inputs. It is one of the two fundamental kinds of supervised prediction.
The prediction can be any value in a range, not just a fixed set of options.
A simple example
Predicting a house’s sale price from its size, location, and features is a regression task — the answer is a number (say, $340,000), not a category. So is forecasting tomorrow’s temperature.
Why it matters
Regression underpins forecasting, pricing, risk scoring, and any task where the goal is to predict a quantity. Alongside classification, it is one of the two core problem types that most practical machine learning falls into.
Related terms
- Classification — predicting a category instead of a number.
- Supervised Learning — the category regression belongs to.
- Feature — the inputs regression uses to predict.
Frequently asked questions
What is regression in machine learning?
Regression predicts a continuous numerical value — such as a price, temperature, or age — based on input features.
How is regression different from classification?
Regression outputs a number on a continuous scale, while classification outputs a discrete category or label.
