The Perceptron as a Roadmap - From Neuron Structure to Artificial Neural Networks

Explore the evolution of artificial neural networks in this graduate-level colloquium lecture by Professor Daniel Alpay. Trace the development from the discovery of neuron structure to modern machine learning algorithms, using the perceptron as a central focus. Delve into the historical context of the perceptron, created by psychologist Frank Rosenblatt in the 1960s for image classification, and its foundation in the McCullogh and Pitts neuron model from 1943. Examine the perceptron algorithm as one of the earliest machine learning techniques and its influence on subsequent artificial neural network structures. Investigate connections to Hopfield networks, associative memories, and function approximation. Discover unexpected links to mathematicians like Agmon, Schoenberg, and Wiener in this comprehensive exploration of the intersection between neuroscience, computer science, and mathematics.

Daniel Alpay: The Perceptron as a Roadmap (Graduate Colloquium in Math, Philosophy and Physics)