The primary artificial intelligence (AI) architecture, which is loosely based on the behavior of neurons in the human brain. All such networks are technically "artificial neural networks" (ANNs) because they are not human.
The neural network is used in image, language and speech recognition, text-to-speech conversion, robotics, diagnosing, forecasting and generative AI, which creates totally unique output. Unlike regular applications that are programmed to deliver precise results (if-then-else), neural networks are "trained" with millions, billions, even trillions of examples of media (text, images, etc.). After the training phase and while doing the processing they were designed for (predict, generate), some neural networks may be able to adapt and improve themselves. See
AI training,
AI model and
generative AI.
A Neuron in a Neural Network
Nothing like "if-then-else" business logic, neural networks cannot be debugged or reverse engineered like regular computer programs. This is a single neuron. See
AI weights and biases.
The Layers
A neural network comprises one input and one ouput layer and any number of hidden layers in between all tied together mathematically. The following network architectures from the Asimov Institute in the Netherlands reveal the variety of network architectures that have been created. For a neural network example that everyone can relate to, see
convolutional neural network.
Neural Network Architectures
AI networks are one of the most researched areas of computing in the 21st century. The examples above from the Asimov Institute in the Netherlands reveal the variety of network architectures that have been created. (Images courtesy of Fjodor van Veen and Stefan Leijnen (2019). The Neural Network Zoo.)