讲座题目：Neural Systems Integrating and Segregating Information from Different Channels 主讲人：K. Y. Michael Wong 教授 主持人：刘宗华 开始时间💁🏼：2019-12-19 10:00:00 讲座地址：闵行校区物理楼314会议室 主办单位：物理与电子科学学院
报告人简介： K. Y. Michael Wong，香港科技大学物理系教授🤽♀️。1978 年获香港大学物理系物理学学士🚴♀️、1981年与1986年分别获美国 UCLA 物理学硕士与博士学位。1986至1989年在帝国大学做博士后🟩🗿，1989至1991年在牛津大学做博士后✸，1982年至1997年在香港科技大学物理系任讲师，1998年至2006年任副教授，2006年至现在任教授🚰。2002年至2006年任香港物理学会主席，现为Journal of Statistical Mechanics: Theory and Experiment编辑(2007-present) ，曾任Special Issue of Philosophical Magazine on David Sherrington’s 70th Birthday (2011) 与Research Topic on Neural Information Processing with Dynamical Synapses, Frontiers in Computational Neuroscience (2012-2013) 的Guest Editor，并获EPL Distinguished Referee 2014。到目前为止已在美国科学院期刊PNAS及物理学顶级期刊PRL等发表论文184篇👷🏽。研究领域为统计物理、相变理论、大脑认知科学、及复杂网络科学。
报告内容🧡： Neural systems gather information from different channels resulting in enhanced reliability. The optimal estimate is given by Bayes' rule, and remarkably, experiments on macaques showed that the brain can achieve this optimum. It is therefore interesting to consider the neural architecture and mechanism underlying this feat. Nevertheless, this leads to some interesting information issues to be addressed. For channel inputs that are correlated in their prior distribution, where and how is the information stored? Furthermore, when the channel inputs are partially correlated, how can the system perform partial integration? Besides performing integration when the channel inputs are correlated, how can the system perform segregation when they are not? Finally, since input integration implies that the system extracts the information common to the channels, can the information of the individual channels be recovered? To answer these questions, I will discuss the results of our study on a decentralized network architecture where same-channel and cross-channel information are processed in parallel. I will provide an explanation of the role played by opposite neurons in the VIP and MSTd areas of the brain, and report a striking discovery that the direct and indirect cross-channel pathways are opposite to each other -- an apparently redundant architecture. |
