Neural Networks: Logic and Applications

This lecture covers the logic of neuronal function, starting with McCulloch & Pitts' work in 1943, followed by Hodgkin & Huxley's model in 1952. It explores the Perceptron model by Rosenblatt in 1958 and the application of neural networks in deep learning by LeCun, Bengio, & Hinton in 2015. The lecture delves into the structure and function of neural networks, including activation functions and their real-world applications. It also discusses the levels of abstraction in neural models, from detailed biological insights to abstracted models with limited biological insight. The exercise involves modeling neuronal function and exploring the STG anatomy, circuitry, and receptive fields.