Neuron

As explained in Data Without Labels

A neuron is a computational unit in a neural network that combines inputs and applies an activation function to produce an output.

A neuron is a small computational unit in a neural network. It receives numerical inputs—either raw data or outputs from other neurons—and produces an output that can feed the next neuron. An artificial neuron is a simplified model of a biological neuron; a perceptron is a mathematical model of this unit. section 8.2.2

To produce its output, a perceptron multiplies each input by a weight, combines the weighted inputs, and passes the result through an activation function. Weights determine how strongly inputs contribute; bias terms also play a role in the perceptron. Activation functions provide the nonlinear step. section 8.2.2 section 8.13

Figure 8.3 A neuron gets the inputs, processes them using mathematical functions, and then generates the output.

Follow the inputs through the neuron's calculation to its output. section 8.2.2

Neurons are commonly arranged in layers. An input layer receives data, hidden layers process it and learn features, and an output layer makes a prediction. In the layered arrangement shown here, a neuron's inputs come from the preceding layer and its outputs go to the next. section 8.2.3

Recognizing a face

A face-identification network trains on images paired with names. As image data passes through its neurons, hidden layers learn attributes useful for distinguishing faces; the output layer produces the prediction. A convolutional neural network uses a more localized connection pattern: each neuron in a convolutional layer connects to only a small region of the input. section 8.2.1 section 8.2.3 section 8.6.1

How Data Without Labels covers it

How deep it goes

Explains numerical inputs, weights, weighted sums, activation, and layered connections; does not work through a numerical calculation.

Examples

Illustrates neurons with a face-identification network, layer diagrams, and localized connections in image classification.

What it assumes

Introduces the unit through a biological analogy and numerical inputs before describing its calculation.

What you can do after

Evaluate/understand it: trace how a neuron transforms inputs and passes its output toward a network prediction.

How it connects

Connects neurons to perceptrons, weights, biases, activation functions, network layers, and convolutional neural networks.

liveBookmanning.com

Read Data Without Labels on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Natural Language Processing in Action, Second EditionAn artificial neuron is a parameterized computational unit that combines inputs to produce an output, typically through an activation function.
• 

  liveBookmanning.com
  Classic Computer Science Problems in Swift: Essential techniques for practicing programmersA neuron is a computational unit that combines input signals with weights, applies an activation function, and produces an output signal.
• 

  liveBookmanning.com
  Grokking AI Algorithms, Second EditionA neural-network node is a computational unit that combines weighted signals and transforms them into an activation.
• 

  liveBookmanning.com
  Optimization Algorithms: AI techniques for design, planning, and control problemsAn artificial neuron is a computational unit that combines weighted inputs and a bias before applying an activation function to produce an output.
• 

  liveBookmanning.com
  Sutskever's ListA computational unit in a neural network that combines weighted inputs and applies an activation function.

Compare how 6 books explain neuron →

Related concepts

• neural network layerpart of · A neural network layer is a computational stage or component that transforms input representations into outputs within a neural network.
• language modelpart of · A language model is a computational model that represents patterns in linguistic tokens to estimate sequence probabilities and predict or generate language from context.
• activation functionuses · An activation function maps a neural-network unit’s or layer’s input to an output, often introducing nonlinearity into the network’s computation.
• perceptronincludes · A perceptron is a single-layer computational model and linear classifier that combines weighted inputs with a bias and applies a threshold or activation rule.
• deep learningpart of · Deep learning is a machine-learning approach that uses neural networks with multiple computational layers to learn representations from data.
• neuroncontrasts with · A biological neuron is a nervous-system cell that receives, integrates, and transmits signals through electrical and chemical activity.

This explainer was generated by Manning's knowledge graph from the text of Data Without Labels. The section links point to the parts of the book it draws on.