• Addresses the fundamental scientific question of how the human brain still outperforms the best computer algorithms in most domains of sensory, motor and cognitive function, as well as the parallel and distributed nature of neural processing (as opposed to the serial organization of computer architectures/algorithms) required to answer it. Explores the biologically plausible computational mechanisms and principles that underlie neural computing, such as competitive and unsupervised learning rules, attractor networks, self-organizing feature maps, content-addressable memory, expansion recoding, the stability-plasticity dilemma, the role of lateral and top-down feedback in neural systems, the role of noise in neural computing. Students taking graduate version complete additional assignments.

• Textbook Information

There is no information available for this course.