Description:

Title: Algorithms of the Early Olfactory System

Speaker: Thomas Cleland, Department of Psychology, Cornell University

Abstract:The network architecture of the olfactory system is peculiar, but well adapted to the particular challenges of chemosensation.  Olfactory signals are intrinsically high-dimensional, lacking the two-dimensional structure of visual images that facilitates their recognition. Odor signal identification is further challenged by background odorants that disrupt the afferent activity patterns on which odor recognition ostensibly depends.  I here illustrate the nature of the problem, describe some computational motifs in olfactory bulb circuits, and present a current theoretical framework for odor learning in the olfactory bulb that promises to enable these difficult identifications under noise.  Moreover, this framework offers a potentially powerful solution to the “small n, large p” problem for classification of relatively unstructured arbitrary data such as medical diagnostics. 

Co-Sponsored by Northwestern University Quantitative Biology Research Training Program.

-----

To subscribe to the Applied Mathematics Colloquia List send a message to seminar-join@esam.northwestern.edu