Neural Networks · Level 2 of 5
Neural Network
A parameterized model built from connected computational units or layers.
Units transform and combine signals; biological neurons are only a loose inspiration.
Example
A network maps image pixels to class scores.
Listen to the definition and example
Audio transcript
Neural Network. A parameterized model built from connected computational units or layers. Units transform and combine signals; biological neurons are only a loose inspiration. For example: A network maps image pixels to class scores.
Explore this concept
A useful analogy
A series of adjustable mixers and transformations that convert an input into a prediction.
Why it matters
This is part of the vocabulary used to read network diagrams and understand parameter updates.
Technical detail
An MLP composes affine maps and nonlinear activations, such as h = σ(Wx+b). Parameters are typically learned through gradient-based optimization.
Common misconception
Artificial neurons are simplified computations, not faithful replicas of biological cells.
Start with
Related concepts
Quick recall question
Try answering before looking back at the definition.