AI Rookies

Backpropagation

Fact

A training method that sends errors backward to update a model’s parameters.

In Plain Words

Backprop is a coach with instant replay. After a missed shot, it fixes the wrist, then the elbow.

It tells a neural network which parts caused the mistake. It is a core step for training deep models.

Related Concepts

Neural-network
Backprop sends the error backward through the Neural-network and helps change its weights.

SGD
Backprop finds which way to change, then SGD updates the parameters.

Deep Learning
Backprop makes many-layer training possible in Deep Learning.

Parameter
Backprop tells each Parameter how it should change.