Dongzhuo Zhou, Professor, Shanghai Jiao Tong University

Haijun Yu, Professor

Dynamics and information coding in balanced neuronal networks

10:00-11:00 August 4 (Friday), Z301

The idea of a balanced state is that the excitatory and inhibitory components of inputs nearly cancel each other, and the neural firing activity is driven by strong fluctuations that intermittently interrupt this cancellation and give rise to the irregularity of neural spikes. Consistent with the hypothesized scenario, balanced synaptic inputs have been observed in experiment, and it is believed that the balanced state is closely related to the efficient coding in real neuronal networks. Previous theoretical and computational works proposed a mechanism to understand the emergence of the balanced state in homogeneous networks and have shown that small perturbations of the balanced state in a network with binary neurons grow exponentially, indicating the chaotic nature of the balanced activity. However, there are several issues that remain to be clarified. First, whether the chaotic dynamics is the underlying mechanism for the irregularity of neural activity in a balanced network. Second, whether a balanced state can exist in heterogeneous networks and how it differs from the balanced state in homogeneous networks. Third, how the balanced state is related to information coding in networks as well as network structure. We will address these issues through modeling, analysis and simulations in this talk.

报告人简介: 周栋焯，教授，博士生导师，研究领域为计算与应用数学，具体方向是计算神经科学，2002年和2007年分别获得北京大学学士和博士，2007-2009年美国纽约大学库朗研究所博士后，2010年进入上海交通大学自然科学研究院/数学科学学院工作，至2015年任特别研究员，2016年至今任教授。现任CNS计算神经科学学会秘书长，CSIAM数学生命科学分会常务理事，获得国家自然科学基金委优青、杰青项目，上海市科委青年科技启明星等，在国际学术刊物 CPAM，PNAS，PRL等发表论文60余篇。