Skip to content
BENIBIX Learn

Layer (Neural Network)

AI📈 Machine LearningNeural Networks

Also known as Hidden Layer, Input Layer, Output Layer

One row of neurons. Data goes through each row in turn.

Think of it like

Think of it like one floor in a tall building. Each floor does its job, then sends the work up.

Example

Lucy's photo model has 20 layers. The first ones spot edges and colours. The middle ones spot eyes and wheels. The last one says cat or car.

Why it matters

The number of layers is a big choice. More layers can learn harder things, but they need more data and more time.

Definition

A layer is one row of neurons in a neural network that work side by side. The input layer takes in the data, and the output layer gives the answer. The layers in between are called hidden layers. Each layer turns what it gets into something a bit more useful for the next. Early layers find simple things, like edges, and later ones find whole shapes.
Each column is a layer. The ones between input and output are the hidden layers.

Common mistake

Thinking more layers always helps. Past a point, extra layers make training slow, and the model may just memorise.

Key takeaways

  1. 1
    A row of neurons. They work side by side.
  2. 2
    In, hidden, out. Three kinds of layer.
  3. 3
    Simple to complex. Each builds on the last.

Learn it properly