UNIVERSITY OF HERTFORDSHIRE COMPUTER SCIENCE RESEARCH COLLOQUIUM presents "Machine learning for hypothesis generation in neuroscience" Dr. Dan Goodman (Department of Electrical Engineering Imperial College London) 17 October 2018 (Wednesday) 1 - 2 pm Hatfield, College Lane Campus Seminar Room C152 Everyone is Welcome to Attend Refreshments will be available Abstract: Methods from machine learning and artificial intelligence have made enormous advances lately in solving real world problems. This has led many to suggest that they might be a good source of inspiration for understanding our own brains. In this talk, I'll argue that they can, but that we shouldn't take our inspiration too literally. The brain is not structured like these algorithms, and they make errors that are very unlike those made by humans. Instead, we should use the ability of machine learning methods to handle real world complexity as a way to search for plausible hypotheses for neural function. I will illustrate this approach with two examples of ongoing work on modelling computations in the early auditory system, and on analysing high dimensional cortical recordings. --------------------------------------------------- Hertfordshire Computer Science Research Colloquium http://cs-colloq.cs.herts.ac.uk