Machine learning (ML) is the technique behind most modern AI: instead of programming a computer with explicit rules, you give it many examples and it learns the patterns itself. Show a model thousands of emails labeled spam or not spam, and it learns to recognize spam it has never seen. Deep learning, a type of machine learning using large neural networks, powers today’s AI assistants and image generators.
Part of AI Basics. Last updated September 2026.
The Main Types of Machine Learning
| Type | How It Learns | Example |
|---|---|---|
| Supervised learning | From examples with correct answers (labels) | Predicting house prices from past sales |
| Unsupervised learning | Finds structure in unlabeled data | Grouping customers into segments |
| Reinforcement learning | By trial and error with rewards | Game-playing systems; refining AI assistant behavior |
| Deep learning | Uses large neural networks with many layers | Language models, image recognition |
How a Machine Learning Model Is Built
- Collect data relevant to the problem.
- Train the model by letting it adjust its internal settings to fit the examples.
- Test it on data it has not seen to check accuracy.
- Deploy it to make predictions on new data.
- Monitor it, because accuracy can drift as the real world changes.
Everyday Business Uses
- Forecasting demand and sales.
- Detecting fraud and unusual activity.
- Recommending products and content.
- Categorizing transactions and support tickets.
- Predicting equipment maintenance needs.
Limits to Keep in Mind
- Garbage in, garbage out: biased or poor data produces biased or poor predictions.
- Correlation is not causation: models find patterns, not reasons.
- Change breaks models: a model trained on last year’s behavior may miss new trends.
Frequently Asked Questions
What is the difference between AI and machine learning?
AI is the broad field; machine learning is the most common method used to build AI systems today.
Do I need to learn machine learning to use AI?
No. Using AI tools requires no ML knowledge. Learning ML matters if you want to build or evaluate models.
What is a neural network?
A model loosely inspired by the brain, made of layers of connected units whose connection strengths are adjusted during training.