What is it?
Artificial neural networks are a software approximation of a biological human brain, which tries to mimic the process which neurons communicate to each other to process information. ANNs are at the very core of Deep Learning, and provide much more versatile and flexible models than traditional Machine Learning modelling.
How does it work?
Proposed by Warren McCulloch and Walter Pitts in the paper “A Logical Calculus of Ideas Immanent in Nervous Activity”, the biological nerve cell served as the basis for the creation of the artificial neuron structure, which is basically a gate which receives one or more binary input and outputs a single binary. These artificial neurons can compute any logical proposition :
With this in mind, Frank Rosenblatt invented the Perceptron, which is slight different than the above artificial neurons, as it implements the threshold logic unit (TLU).