主讲人：汪小京教授

The past, presentand future of Computational Neuroscience and AI

计算神经科学和人工智能的历史、现状和发展前景

Speaker:Prof. Xiao-Jing Wang, New York University

Inviter:Prof. Tian Chi

Time:13:30-14:45, Mar 9, 2018

Venue:230 Haike Road, Auditorium, L Building, ShanghaiTech 海科路230号（海科路科苑路）生命学院 L楼 一楼报告厅

Global Professor of Neural Science, co-director of the Swartz Center for Theoretical Neuroscience, Adjunct Professor of Physics and Mathematics at New York University.

2012-2017, Founding Provost and Vice President for Research at NYU Shanghai.

Prior to 2012, Professor of Neurobiology at Yale University.

教授简介：汪教授是理论与计算神经科学专家，研究重点是认知功能的脑机制，尤其以在短期记忆的细胞基础、决策的神经机制、大脑细胞网内信息交流与同步、抑制神经元功能等领域的研究而著称。他的团队开创了被称为“大脑CEO”的前额皮质神经网络模型研究的前沿。近年来，汪教授的团队建立了大型脑神经网络的神经生物学仿真模型，深入研究由认知功能及控制的灵活性行为，及其在人工智能和精神医学的应用。

Profile:Wang is an expert in Theoretical and Computational Neuroscience, with a special interest in the neurobiology of executive and cognitive functions. His group has pioneered neural circuit theory of the prefrontal cortex, which is often called the “CEO of the brain”. In particular, Wang is known for his work on the cellular basis of short-term memory, neural mechanisms for decision-making, communication, and synchronization through inhibitory neurons in the brain. His research group has also pioneered biologically-realistic large-scale circuit models of the primate cortex, with the goal to elucidate the complex global brain mechanisms of cognitive functions flexible behavior as well as applications to artificial intelligence and Psychiatry.

讲座摘要：类脑智能和脑技术（BrainTech）的发展在全球范围得到了社会和企业的高度重视。跨越式的类脑技术创新需要坚实的理论基础。就拿现在最热门的“深度学习”来说，其理论来源于90年代的研究突破。神经科学对大脑视觉系统的研究，产生了“深度网络”的概念，人们为此建立数学模型，再有Geoff Hinton, Yann LeCun 等人发展的训练网络学习的方法，才有了今天的深度学习框架。从实验神经生物学到人工智能，需要用数学模型来建立一座桥梁， 计算神经科学就是这座桥梁。而计算神经科学不仅仅是一座桥梁，它在脑科学中扮演着理论先导的角色。这是21世纪的跨学科领域，越来越吸引数理化领域的杰出年轻人。我将谈谈计算神经科学和人工智能的历史、现状和发展前景。

Seminar Abstract:Artificial intelligence has recently generated tremendous excitement in the society, the advances are a result of basic research over the last decades. In particular, deep learning theory was formulated in the 1990's, inspired by neuroscientific studies of the visual system from the 1960's to the present time. A mathematical link is required to translate neural circuits in the brain (wet matter in a biological organ) into models and algorithms. Theoretical or Computational Neuroscience is the field that serves precisely such a bridge. Furthermore, just like Theoretical Physics in Physics, Computational Neuroscience plays a fundamental role in Neuroscience for making sense of empirical data and discovering general principles. This highly cross-disciplinary field is on the verge to take off in China, attracting young talents from Physics, Mathematics, Engineering and Computer Science. In this talk, I will introduce Computational Neuroscience and discuss its interplay with AI in the future.

欢迎全校师生参加！

上海科技大学生命科学与技术学院