What is supervised learning?
Length:
3 min
Published:
June 9, 2026
What supervised learning means
Supervised learning is the most common form of machine learning. You train a model on a dataset where every example comes with the correct answer attached, called a label. The model studies thousands of input-and-answer pairs and learns the pattern that connects them. Once trained, it can predict the answer for new inputs it has never seen.
A spam filter is a classic example. You feed it emails that people have already marked as "spam" or "not spam". The model learns which words and patterns point to spam, and then sorts incoming mail on its own.
In plain words
Think of teaching a child to recognise fruit. You hold up an apple and say "apple", then a banana and say "banana", over and over. After enough examples, the child names a fruit you never showed them. Supervised learning works the same way: the labels are your answers, and the model learns to generalise from them.
When to use it
- You have labeled data. Past examples already tagged with the right outcome, like approved or rejected loan applications.
- The answer is well defined. You want to predict a category (classification) or a number (regression), such as a price or a probability.
- The task repeats. Sorting tickets, detecting fraud, forecasting demand, scoring leads, all benefit from a model that learns from history.
The two other big families are unsupervised learning, which finds structure in data without labels, and reinforcement learning, which learns from rewards instead of correct answers.
Common pitfalls
- Bad labels poison the model. If your training answers are wrong or inconsistent, the model learns the mistakes. The quality of your labels caps the quality of your predictions.
- Overfitting. A model can memorise the training set instead of learning the pattern, then fail on real data. Always test it on examples it never saw during training.
- Biased data, biased model. If the historical data favours one group, the model will too. Check what your labels actually represent before you trust them.
Related articles:
- What is machine learning? - The broader field that supervised learning belongs to.
- What is reinforcement learning? - Learning from rewards instead of labeled answers.
- What is an AI model? - What you actually get once training is done.
Want to stay one step ahead?
Don't miss our best insights. No spam, just practical analyses, invitations to exclusive events, and podcast summaries delivered straight to your inbox.