A model built from layers of connected nodes whose weighted links are adjusted during training until the network maps inputs to the right outputs.
Neural Networks, loosely inspired by biological neurons, are the basis of deep learning. Each one is built from layers of small mathematical functions, called neurons by analogy, and every layer passes its output to the next through weighted connections. During training, the weights of these connections are adjusted so the network learns complex patterns in the data.
See also Deep Learning, Private AI, Digital Transformation and Enterprise Architecture.