Neural Network

A neural network is a stack of simple math units that pass signals forward to make a guess.

What it does

A neural network is a stack of simple math units that pass signals forward to make a guess.

How it works

• Take input numbers.
• Pass them through connected units called neurons.
• Use the final signal as the answer.

Uses and applications

• A phone can recognize a face in a photo.
• A game AI can choose a move from screen information.
• A speech app can help turn sound into words.

In short

Neural networks are the learning part inside methods such as DQN.

Related methods

• Deep Learning: Stacks many more layers of the same building blocks.
• Logistic Regression: Equivalent to one neuron with a sigmoid output.
• RNN (Recurrent Net): Adds memory so the network can read sequences.

Back to Neural networks and deep learning · All topics

The interactive demonstration and the Python walkthrough for this topic need JavaScript.