Home / Wiki / N / Neural Network

Neural Network

Technology and VendorsUpdated 16 Nov 2026

Neural network is the everyday name for an artificial neural network, a computing model made of layers of simple connected units whose connection weights are adjusted during training until the model maps inputs to useful outputs. The main entry explains the structure, training methods and main types.

In computing and finance the short form almost always means the artificial kind. The longer name is used mainly to separate it from biological networks of neurons, which inspired the idea but work very differently. Networks with many layers are the basis of deep learning, and the same family of models underlies image recognition, speech systems and the large language models behind modern chat assistants.

Banks and trading firms use neural networks within machine learning systems for fraud detection, credit scoring, document reading and forecasting. They are also a major theme for investors, since training and running large networks drives demand for specialized chips, data centres and electricity. Their main drawbacks are the large amounts of data they need and the difficulty of explaining individual decisions to customers and regulators.

Last updated 16 Nov 2026