What is Supervised Learning?

What is Supervised Learning?

Supervised learning is a machine learning paradigm in which a model is trained on labeled data — input-output pairs where the correct answer is known. The model learns to map inputs to outputs and can then make predictions on new, unseen data.

Why It Matters

Supervised learning is the most widely used form of machine learning in practice. Classification (is this email spam?), regression (what will this house sell for?), and even the supervised fine-tuning step of LLM training all use supervised learning. It's the paradigm most people encounter first and the backbone of countless production AI systems.

How It Works

1. Labeled dataset — each training example has an input and a corresponding label (the "ground truth").
2. Model selection — choose an algorithm (linear regression, decision tree, neural network, etc.).
3. Training — the model processes training examples, makes predictions, computes error using a loss function, and adjusts its parameters via gradient descent to reduce error.
4. Evaluation — test on held-out data to measure accuracy, precision, recall, or other metrics.
5. Prediction — apply the trained model to new inputs.

Two main tasks:

• Classification — predict a category (spam/not spam, cat/dog, positive/negative sentiment)
• Regression — predict a continuous value (price, temperature, probability)

Example

A bank trains a fraud detection model on millions of past transactions, each labeled "fraudulent" or "legitimate." The model learns patterns (unusual amounts, foreign locations, rapid succession) and can flag suspicious new transactions in real time.