Machine Learning
Discover how computers learn from examples, get trained, and make predictions.
1. Learning from examples
In machine learning, programmers don't write every rule. Instead, they give the computer many examples. After seeing thousands of labeled cat and dog photos, it can figure out the differences on its own.
2. Training
Feeding examples to an A.I. so it can learn is called training. Training can take hours or even months and uses a lot of computing power for large systems.
3. The model
The result of training is a model — a system that stores the patterns it learned. Think of it as the A.I.'s "knowledge."
4. Making predictions
A trained model makes predictions about new data it has never seen. A model trained on weather history can predict tomorrow's weather. Predictions aren't always correct, but good models are right most of the time.
Field Vocabulary
A type of A.I. where computers learn rules from examples instead of being programmed with every rule.
The process of feeding examples to an A.I. so it can learn patterns.
The result of training — a system that uses learned patterns to make decisions or predictions.
An output where the A.I. estimates an answer based on patterns it learned.