Machine Learning for Brain Data
How AI learns from examples
Supervised learning
In supervised learning, the model sees input data and the correct answer. For example, it may learn to map a brain pattern to a chosen letter, movement, or intention.
Classification and regression
Classification predicts a category, such as left or right. Regression predicts a number, such as the strength of a movement intention or a cursor position.
Training and testing
A model is trained on one set of examples and tested on different examples. Testing helps us estimate whether the model will work on new brain data, not just the data it memorized.
Beginner insight
If a model performs well on training data but poorly on new data, it may be overfitting. That means it learned the examples too specifically.
What does supervised learning require?
Supervised learning uses labeled examples to teach a model.
Correct answer: Input data and correct answers
What is overfitting?
An overfit model may fail to generalize to new data.
Correct answer: When a model learns training examples too specifically