Restricted Boltzmann Machine (RBM)

A Restricted Boltzmann Machine (RBM) is a type of neural network designed for learning patterns in data. It consists of two layers: visible units (which represent input data) and hidden units (which capture underlying features). Connections only exist between these layers, not within a layer, enabling efficient learning of complex structures. During training, the RBM adjusts relationships between visible and hidden units to model data distribution, allowing it to generate or recognize similar patterns. It's often used for feature extraction, dimensionality reduction, or as a building block for deep learning systems.