Neuron
A neuron is a computational unit in a neural network that transforms inputs, typically using learned parameters and an activation function, to produce an output.
Natural Language Processing in Action, Second Edition
An artificial neuron is a parameterized computational unit that combines inputs to produce an output, typically through an activation function.
- How it is used
- Provides the starting unit for building neural networks that make predictions from numerical representations of words and names.
- How deep it goes
- Works through weighted inputs, bias, threshold and logistic activation, then implements and trains a single-output PyTorch model.
- Examples
- Uses a Python word-count calculation and a PyTorch classifier trained on character n-grams from first names.
- What you can do after
- Build it from scratch as a weighted-sum threshold unit, then apply it via PyTorch to binary classification.
Read how Natural Language Processing in Action, Second Edition explains neuron →
Classic Computer Science Problems in Swift: Essential techniques for practicing programmers
A neuron is a computational unit that combines input signals with weights, applies an activation function, and produces an output signal.
- How it is used
- Models the weighted signal-processing unit used to build and train a feed-forward classification network.
- How deep it goes
- Defines the computation and implements neuron state and training support, while deferring dot-product fundamentals and calculus derivations.
- Examples
- Uses Swift neuron and layer code plus an iris classifier with four inputs, six hidden neurons, and three outputs.
- What you can do after
- Build it from scratch as a Swift neuron that produces activated outputs and retains state for backpropagation.
Grokking AI Algorithms, Second Edition
A neural-network node is a computational unit that combines weighted signals and transforms them into an activation.
- How it is used
- Uses neurons as connected processing units that turn input features into an artificial neural network’s classification.
- How deep it goes
- Works through weights, sums, activations, and forward propagation; treats weight learning separately from the pretrained prediction walkthrough.
- Examples
- Traces collision prediction from driving-condition features and illustrates layer calculations with a Python neural-network listing.
- What you can do after
- Build it from scratch as a small network that combines weighted inputs and applies activations to make a prediction.
Read how Grokking AI Algorithms, Second Edition explains neuron →
Optimization Algorithms: AI techniques for design, planning, and control problems
An artificial neuron is a computational unit that combines weighted inputs and a bias before applying an activation function to produce an output.
- How it is used
- Explains the basic computational unit used to construct neural networks for classification and other machine-learning mappings.
- How deep it goes
- Defines weighted inputs, bias, threshold, and output, then illustrates layer computation without developing neuron-level training mathematics.
- Examples
- Uses vehicle-image object recognition and a Python penguins classifier with four inputs, ten hidden neurons, and three output classes.
- What you can do after
- Evaluate/understand how weighted inputs and biases become layer activations in a small neural-network classifier.
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.
- How it is used
- Uses neurons as the input-processing units that pass information through a neural network toward a prediction.
- 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 you can do after
- Evaluate/understand it: trace how a neuron transforms inputs and passes its output toward a network prediction.
Sutskever's List
A computational unit in a neural network that combines weighted inputs and applies an activation function.
- How it is used
- Defines the computational unit used to describe AlexNet’s network architecture and its biological framing.
- How deep it goes
- Explains weighted inputs, optional bias, and activation, then notes why counts become ambiguous in convolutional layers.
- Examples
- Uses AlexNet’s roughly 650,000 neurons and 60 million parameters to contrast two architectural counts.
- What you can do after
- Evaluate neuron counts as architectural descriptors and distinguish them from parameter counts when comparing models.
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 transforms a neural-network unit’s input into its output, typically introducing nonlinearity that increases the network’s representational capacity.
- 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.
- neural networkpart of · A neural network is a computational model of interconnected, parameterized units that learns transformations from inputs to outputs.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.