CMPS 535B Neural Networks

CMPS 535. NEURAL NETWORKS (Credit, 3 hours). This course will consider design, architecture, and implementation of neural networks. Neural networks are becoming increasingly versatile due to their ability to solve difficult nonlinear problems that are not solvable using traditional methods. Inherently parallel design and ability to interact with the environment make neural networks ideal for large applications. Topics include neural networks as emerging technology, perceptions, associative memory networks, radial-basis networks, spline networks, recurrent networks, neural learning, gradient descent method, and back-propagation. Issues related to neuro-computing hardware and neuro-VLSI implementation will be discussed. Neural networks will be examined as problem solving tools as compared with fuzzy systems and expert systems. (Prerequisite: Consent of instructor.)

Credits

3

Cross Listed Courses