Neural Network

Also known as · neural networks

Layers of connected units (neurons) that transform inputs into outputs through learned weights. Each neuron is a specialist feature detector, and the network learns its own features.

When to use

Use it to explain why deep learning beats feature-engineered models: on the Derech case an MLP reached 83 percent accuracy against 72 percent for the best earlier model, because it learns feature interactions that trees and linear models miss.

See Lec 8.

Appears in

No references yet.