Designation: Required Course
Course Level: Graduate
Prerequisite(s) by Topic:
COE 580 or Consent of the Instructor
Introduction to neural computation. Biological neurons. Fundamental concepts behind various models of neural networks. Functional equivalence and convergence properties of neural network models. Adaptation and learning in neural networks: associative, competitive, inhibitory, and adaptive resonance models of learning. Back-propagation, Hopfield Nets, Boltzmann machines, Cauchy machines, ART, and feature map (Kohonen model). Cognitron and neocognitron. VLSI, optical, and software implementations. Potentials and limitations of neural networks. Applications to vision, speech, motor control and others. Projects.