Event
Prof. Bing Brunton from the University of Washington- Extracting spatio-temporal coherent structures
Monday, August 1, 2016
12:00 p.m.
AV Williams Building, Room 1146
Pamela Abshire
pabshire@umd.edu
Please join Prof. Pamela Abshire (ECE/ISR) for a lecture with Prof. Bing Brunton from the University of Washington. Prof. Brunton is an Assitant Prof in the Department of Biology and the UW Insitiute of Neruengineering. She is also a Data Science Fellow of the UW eScience Institute and a faculty member of the Graduate Program in Neuroscience.
Abstract:
There is a broad need in neuroscience to understand and visualize large-scale recordings of neural activity, big data acquired by tens or hundreds of electrodes recording dynamic brain activity over minutes to hours. Such datasets are characterized by coherent patterns across both space and time, yet existing modal decomposition techniques are typically restricted to analysis either in space or in time separately. I will talk about recent work adapting dynamic mode decomposition (DMD), an algorithm originally developed for studying fluid physics, to large-scale neural recordings. DMD is a modal decomposition algorithm that describes high-dimensional dynamic data using coupled spatio-temporal modes; we may think of DMD as a rotation of the low-dimensional PCA space such that each basis vector has coherent dynamics. The resulting analysis combines key features of performing PCA in space and power spectral analysis in time, making it particularly suitable for analyzing large-scale neural recordings. In addition, DMD interfaces dynamical systems theory with modern big data measurements of complex systems where it is infeasible to acquire true equations of motion. Data-driven model predictions do not typically hold for long-term; however, even short-term DMD predictions can inform design of neural decoding or feedback controllers.
More information about Prof. Brunton can be found HERE
