Computational & Data Science seminar: An Evolutionary Framework for Developing Models of Cognition
Friday, April 20, 2018 - 12:30pm

What: April 20th Computational & Data Science Seminar


When: 12:30 pm April 20th, 2018


Where: Cramer Hall 418

Title:  An Evolutionary Framework for Developing Models of Cognition 
Abstract In this talk, I present a generalizable framework that employs evolutionary algorithms combined with a GPU-accelerated spiking neural network (SNN) simulator that can be used to rapidly evolve models of cognitive function by virtue of replicated neuron firing patterns. We applied the framework to a dataset recorded from rat retrosplenial cortex (RSC), a brain region known for its involvement in spatial navigation. Our findings show that realistic and complex representations of space emerge in networks evolved to replicate neuronal function, and suggest that this method may be extended to other datasets in order to develop sophisticated and realistic large-scale models of cognitive function. This method offers a novel and efficient way of modeling in which larger networks can be trained using an embedded subset of seed neurons, and can work rapidly despite the need for both training and testing of the networks, with convergence of the evolutionary algorithm occurring within 20 generations in this case.
Speaker: Emily Rounds earned a M.S. degree in Cognitive Neuroscience and a Ph.D. in Psychology and concentration in Cognitive Neuroscience at the University of California, Irvine in the Cognitive Anteater Robotics Laboratory of Dr. Jeffrey Krichmar. There she studied learning in the context of spatial navigation and episodic memory, with her doctoral work focusing on the development of an evolutionary framework for replicating electrophysiologically observed neuronal firing patterns in spiking neural networks. Her research interests are in learning and memory in the context of computational modeling.
Following the talk, at 1:30, in the same room:
Will Garrick will lead a hands-on workshop on using Matlab on PSU's HPC cluster. Interested parties are encouraged to request access to the HPC cluster prior to the session, and to bring their laptops to the event so they can set up everything necessary to begin using cluster resources.
Link to request a PICS account For more info about this workshop, email
For more info about the overall seminar series:
To add this seminar to your calendar:  calendar link
Please forward this announcement to students and colleagues whom you think might be interested in this seminar series. Thanks!

Wayne Wakeland and Bruno Jedynak