Physics of Learning: From Living Systems to Artificial Neural Networks

Learning can be broadly defined as the process by which a system acquires information (or knowledge), retains it, and uses it to improve future performance in specific tasks. In this talk, I will use this general framework to compare and contrast learning mechanisms across a diverse range of systems—from adaptive responses in E. coli chemotaxis, to Hebbian learning in biological neural networks, to stochastic gradient descent in artificial neural networks.