Â鶹´«Ã½

Neural networks are mathematical models of how the brain works. They store knowledge in the strength of the connections between simple processing units meant to model neurons. Cottrell's research is aimed at understanding how the brain works by building working models of cognitive processes, such as face and emotional expression recognition. However, one would not use these programs to recognize Osama Bin Laden in an airport, as the difference between cognitive models and artificial intelligence programs is that a good cognitive model makes the same mistakes people do! Cottrell's models of facial expression recognition have provided a unified treatment of data that has often been used to support competing theories of how humans recognize expressions. Cottrell has also used neural networks in the arcane area of "computational philosophy," which aims to answer philosophical questions via computer simulations. In the long term, his work will have implications for education and the treatment of social and psychological disorders, including autism, in which the ability to recognize and respond to emotions in others is impaired. Cottrell can also speak about how culture influences facial expression and recognition, and about machine learning in general.

Capsule Bio:

Garrison Cottrell joined the UCSD faculty in 1987. He is one of five faculty members running UCSD's Artificial Intelligence Laboratory. He is Director of UCSD's Interdisciplinary PhD program in Cognitive Science, and co-principal investigator in the Perceptual Expertise Network. Cottrell is also the director of the Temporal Dynamics of Learning Center, an NSF-funded Science of Learning Center. He received his Ph.D. from the University of Rochester in 1985.

Selected Publications: