Introduction

Machine Learning enables computers to learn from data without explicit programming. By recognizing patterns and making predictions, ML systems continually improve their performance over time, leading to more accurate and efficient solutions. Artificial Neural Networks (ANNs) are the fundamental building blocks in most of the common ML algorithms, enabling systems to handle sophisticated challenges like image and speech recognition. They are inspired by the structure of the human brain.

ANN is mainly composed of neurons, placed systematically in layers, starting from the input payer. Each neurons pass data from one layer to another, compute data according to the weights distribution and passes results to the output layer. Depending on the errors on the predicted output, a feedback loop is created to reform the weights in the neuron, which are recalculated repeatedly until desired output is achieved. The intricacies lie in the finer details, but this represents the complete process of the ANN. While data pre-processing and output interpretation vary depending on the specific problem at hand, this documentation will primarily concentrate on the core mechanics: how neurons work and how artificial neural networks process information internally.