GoCareerGo Tutorials

Machine Learning Tutorial

Supervised learning, classification, neural networks and core ML algorithms.

Backpropagation

The algorithm neural networks use to learn — computing how much each weight contributed to the error, then nudging weights to reduce it.

Forward pass — compute the prediction and the loss
Backward pass — use the chain rule to compute the gradient of the loss w.r.t. each weight
Update weights using gradient descent (or a variant like Adam)

Tip

One-liner for interviews: 'backprop is just the chain rule applied layer by layer, computed efficiently.'