Neural network (ANN)
Parametric models organized in layers of simple units; strengths are learned from data via optimization.
An artificial neural network stacks layers whose weights encode behavior; training adjusts those weights using labeled data, objectives, or reinforcement signals, typically via gradient descent and backpropagation.
Architectures range from shallow perceptron stacks to Boltzmann machines, deep belief networks, deep convolutional nets in vision, recurrent predecessors in sequence tasks, and today’s dominant Transformers for text and more see also deep learning.