【学术视频】统计物理与神经计算国际研讨会 | 英国阿斯顿大学David Saad教授

更多精彩视频登陆网站

www.koushare.com

图 | David Saad

题   目：Function-Space Entropy in Deep-Learning Networks

报告人：David Saad单   位：Aston University, UK时   间：2019-10-04地   点：中山大学

扫码观看精彩报告视频

报告提纲

• Deep Learning machines
• Statistical mechanics of learning from examples and why entropy in function-space matters
• Continuous and discrete weights, dense and sparse networks-framework and results
• Correlated weights and convolutional neural networks
• Sensitivity to perturbations and finite-size effects
• Summary and future work
  

个人简介

David Saad obtained a BA in Physics and a BSc in Electrical Engineering at the Technion, Haifa, Israel and later on an MSc in Physics (relativistic field theory) and a PhD in Electrical Engineering (neural networks) at the Tel-Aviv University. In 1992 he joined the neural networks group in the physics department at Edinburgh University first as a postdoc and later on as a lecturer, working mainly on theoretical issues. In 1995 he joined the Neural Computing Research Group at Aston as a lecturer and was promoted later on to a reader (1997) and subsequently to a professor (1999).  Between 2006-2012  he had been the Head of the Mathematics Group and again since 2015. His research interest includes:

• Statistical mechanics of disordered systems

• Advanced Inference in complex systems

• Error Correcting Codes 

• Multiuser communication (CDMA, broadcasting) 

• Hard Computational Problems 

• Computing with noise

• Distributed resources in networks (including routing and smart grids)

• Learning from data and neural networks

会议简介

2019年10月4日-6日，统计物理与神经计算国际研讨会由中山大学物理学院主办，这是在该校举办的第一届物理，机器学习与计算神经科学交叉的国际会议，会议邀请了这一领域近年来作出杰出贡献的国内外专家参与讨论，并围绕神经网络的计算建模，理论研究，生物机制的最新进展展开。

—— ——往期精彩回顾—— ——●【学术视频】统计物理与神经计算国际研讨会 | Alexis Dubreuil: Disentangling the roles of dimensionality and cell classes in neural computation●【学术视频】机器学习-计算化学Workshop | 复旦大学刘智攀教授：Global optimization based Neural Network Potential and Reaction Prediction Artificial Intelligence●【学术视频】第一届量子算法与软件国际研讨会 | 中科院计算技术研究所孙晓明研究员：Optimal Space-Depth Trade-Off of CNOT Circuits●【学术视频】第13届全国分子束外延学术会议 | 清华大学郝智彪教授：氮化物纳米线量子点分子束外延与特性研究●【学术视频】纪念黄昆先生诞辰一百周年暨半导体学科发展研讨会 | 中科院上海微系统与信息技术研究所王浩敏研究员：基于石墨烯纳米带的异质结构筑与逻辑器件探索●【学术视频】第五届全国统计物理与复杂系统学术会议 | 北京工业大学刘鑫副教授：流体纽结复杂系综级联退化过程的拓扑学研究●【学术视频】固态量子器件暑期学校与国际学术研讨会 | 清华大学张浩副教授：Majorana nanowires & topological quantum computation From quantized Majorana to topological qubit?●【学术视频】Workshop on Out of Equilibrium Soft Matter | 苏州大学施夏清副教授：Linking alignment-dominated to repulsion-dominated active matter●【学术视频】第十九届全国催化学术会议 | 河北工业大学李敬德教授：Mechanistic Study of Heterogeneous Catalysis System through DFT-based Kinetic Monte Carlo Simulation

END

扫描二维码，关注微信公众号

戳这里，观看精彩视频哟！