Skip to main content
All terms
Foundations

Neural Network

A model made of layers of simple units connected by adjustable weights.

Definition

A neural network is a model made of layers of simple units (neurons) that pass numbers to each other through weighted connections. Each unit adds up its inputs, each scaled by a weight, then bends the result through a simple curve so the network can capture more than straight-line patterns. During training the weights are tuned automatically (by a method called backpropagation) so the network turns inputs into useful outputs. It is the core building block of deep learning.