Learning algorithms are sets of rules, usually expressed using mathematical equations or computer instructions, that enable a system to improve its performance on the basis of its own experience. Also called learning procedures, methods, or rules, learning algorithms are key components of mathematical models of animal learning and of technological devices engineered to improve their behavior as they operate. Learning algorithms have been studied extensively in artificial intelligence, psychology, computational neuroscience, statistics, and engineering. They are used in many applications in science, industry, and finance that require fitting models to or finding patterns in collections of data or that require learning skills needed to solve specific tasks. Learning algorithms play major roles in artificial neural network systems, where they provide rules for adjusting the strengths, or weights, of connections between network elements. These connections correspond to synapses by which neurons communicate with other neurons and with...