liveBook · Manning

Nonlinear mapping

As explained in Applied Reinforcement Learning

Nonlinearity is a mathematical property that allows a model's output to vary in ways that cannot be represented by a single linear transformation.

Stacking weighted sums alone cannot capture complex patterns: any number of linear layers collapses into one linear transformation. A nonlinear activation after each layer lets a network model more complex relationships. ReLU is commonly used in hidden layers, while linear outputs can represent Q-values. section 8.2

Figure 8.5 The building block of deep networks: neurons and activation functions

See how neurons combine weighted inputs with activation functions. section 8.2

How Applied Reinforcement Learning covers it

How deep it goes
Explains layerwise linear transformations and activations, including why activations prevent stacked layers collapsing into one linear transformation.
Examples
Uses ReLU hidden-layer activations and linear outputs for Q-values as a practical deep reinforcement learning configuration.
What you can do after
Understand how activation functions let deep networks represent relationships beyond a single linear transformation.
How it connects
Connects nonlinear mapping to activation functions, deep neural networks, feature learning, and Q-value estimation.

Also explained in

Compare how 11 books explain nonlinear mapping →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Applied Reinforcement Learning. The section links point to the parts of the book it draws on.