12.3 Neural Networks

A neural network connects many artificial neurons into layers: an input layer that receives the data, one or more hidden layers that transform it, and an output layer that gives the final answer. Each connection has its own weight, and the network learns by adjusting these weights.

Consider an Ethiopian bank scoring loan applications. The input layer reads income, age, and repayment history; hidden layers combine these into risk patterns; the output layer produces a single approval probability.