分论坛 > 长沙 > 新闻动态
CCF YOCSEF长沙西安武汉东北大区四地联动将于2018年8月18日举行人工智能研究前沿及创新应用报告会
2018-07-29 阅读量:1365 小字

中国计算机学会青年计算机科技论坛

CCF Young Computer Scientists & Engineers Forum

CCF YOCSEF

长沙 西安 武汉 东北大区

2018818(星期六) 8:30-12:00 举行

人工智能研究前沿及创新应用报告会

敬请光临


      近年来,随着人工智能技术的发展,特别是像围棋机器人、无人驾驶汽车等人工智能应用的出现,人们目睹了人工智能在学术界、工业界和政府部门的各种激动人心的应用。为了跟踪人工智能的研究前沿和创新应用,例如作为人工智能一个重要分支的计算智能技术,借助自然界(生物界)规律的启示,设计出求解问题的算法,因为具有自学习、自适应的特征和简单、通用、强、适于并行处理的优点,计算智能技术已经在大数据处理、机器人、图像识别、工业生产、金融、航空航天、联想记忆、模式识别、知识自动获取等各领域得到广泛的应用。长沙、西安、武汉、东北大区首次四地联动,与湖南大学信息科学与工程学院联合承办人工智能研究前沿及创新应用报告会,邀请了多位人工智能领域的知名专家,分享人工智能领域的最新研究成果和行业应用,围绕一些共同关注的话题和热点问题进行广泛讨论。


8:00-8:20

8:20-8:30 嘉宾致辞

第一阶段 特邀报告

8:30-9:10     金耀初(Yaochu Jin  英国萨里大学教授IEEE Fellow长江学者

                                          IEEE Transactions on Cognitive and Developmental Systems主编

  Complex & Intelligent Systems主编

 报告题目:Data-driven surrogate-assisted evolutionary optimization of expensive optimization problems


9:10-9:50      Kay Chen Tan       香港城市大学教授, IEEE Fellow

    IEEE Transactions on Evolutionary Computation主编

                    报告题目:Differential Evolution-based Methods for Numerical Optimization


9:50-10:30   何海波  美国罗德岛大学教授, IEEE Fellow

                                  IEEE Transactions on Neural Networks and Learning Systems主编

                    报告题目 AlphaGo Zero and Beyond: Integrated Learning and Control for Machine Intelligence


10:30-10:50 合影/茶歇

10:50-11:30 Qingfu Zhang  香港城市大学教授, IEEE Fellow,长江讲座教授

报告题目:关于进化计算与传统优化关系的思考


 11:30-12:00 公茂果 西安电子科技大学教授、国家优秀青年科学基金获得者、YOCSEF西安主席

                 报告题目:深度神经网络的多目标演化学习


执行主席: 湖南大学信息科学与工程学院副教授,YOCSEF长沙主席

执行主席:   湖南大学信息科学与工程学院副教授,YOCSEF长沙学术委员


特邀讲者:金耀初

金

金耀初 (Yaochu Jin)是英国萨里大学计算科学系计算智能首席教授, “自然计算与应用研究组主任,萨里大学数学与计算生物学中心共同负责人。IEEE FellowIEEE Distinguished Lecturer,教育部长江学者奖励计划讲座教授,芬兰国家技术创新局芬兰讲座教授"。分别于198819911996年在浙江大学电机系获学士、硕士及博士学位,并于2001年在德国波鸿鲁尔大学神经信息研究所获工学博士学位。金耀初教授主要研究领域为进化优化、进化多目标学习、进化发育系统和自组织系统。目前担任IEEE Transactions on Cognitive and Developmental Systems主编,Complex & Intelligent Systems主编。

报告摘要: This talk discusses the main challenges in data-driven surrogate-assisted evolutionary optimization of expensive problems. Fundamental issues such as surrogate model selection, surrogate model management and model training using advanced machine learning techniques in single and multi- and many-objective optimization will be discussed. Challenges in handing sparse data or big data and recent advances in surrogate-assisted optimization of high-dimensional expensive optimization problems will be presented. Finally, real-world examples of data-driven aerodynamic optimization of a ventilation system and a full-scenario airfoil optimization will be given.


特邀讲者:Kay Chen TAN

Kan

Kay Chen Tan is a full Professor with the Department of Computer Science, City University of Hong Kong, Hong Kong. He is the Editor-in-Chief of IEEE Transactions on Evolutionary Computation (IF:10.629), was the EiC of IEEE Computational Intelligence Magazine (2010-2013), and currently serves on the Editorial Board member of 10+ international journals. He is an elected member of IEEE CIS AdCom (2017-2019) and was an IEEE Distinguished Lecturer (2015-2017). He has published 200+ refereed articles and 5+ books. He is a Fellow of IEEE.


报告摘要:

Differential EvolutionDEis arguably one of the powerful metaheuristics for solving numerical optimization problems. Although considerable research has been devoted to the development and improvement of DE, there exist several open issues. This talk will discuss our recent works on designing new DE operators and algorithms to overcome limitations of existing approaches in handling single and multi-objective optimization problems. Application of the proposed DE for solving difficult minimax optimization problems will also be presented.


特邀讲者:何海波

何海波

何海波博士是美国罗德岛大学(University of Rhode Island)的讲席教授(Robert Haas Endowed Chair Professor),智能计算与自适应系统实验室主任,IEEE Fellow

何海波主要从事智能计算以及其在智能电网,大数据,深度学习,机器人应用等方向的研究。已出版学术著作1本,编著1, 编著会议论文集6本,在权威学术期刊和会议上发表论文300多篇。其发表的论文在专业领域产生了深远的影响,包括IEEE Transactions on Knowledge and Data Engineering上高引用论文(单篇论文引用超过3000),多篇进入Essential Science Indicators ESI)高引用论文,IEEE Trans. Information Forensics and Security封面论文,IEEE 通信协会最佳阅读论文(IEEE Communications Society Best Readings 等等。

      何海波教授目前担任期刊IEEE Transactions on Neural Networks and Learning Systems(影响因子:7.982)的主编 Editor-in-Chief)。曾担任10多个IEEE 各类技术委员会主席和副主席,包括IEEE智能计算协会新兴技术委员会主席(Chair, IEEE CIS Emergent Technology Technical Committee)IEEE智能计算协会神经网络技术委员会主席(Chair, IEEE CIS Neural Network Technical Committee), IEEE智能计算智能电网副主席(Vice Chair ,IEEE CIS Smart Grid Task Force)等等。何海波教授曾获得2014IEEE国际通信会议最佳论文奖(IEEE ICC 2014 Best Paper Award),2014IEEE智能计算协会杰出早期职业发展成就奖(IEEE CIS Outstanding Early Career Award), 2011年美国自然科学基金委杰出青年奖(National Science Foundation CAREER Award), 2011年普罗维登斯商报评选的罗德岛州创新明星奖(Providence Business News PBN "Rising Star Innovator" Award)

报告摘要:

The recently advancements in artificial intelligence, especially the mastering of the Go game from Google AlphaGo/AlphaGo Zero, has witnessed tremendous excitements worldwide from academia, industry, and government. This impressive progress not only demonstrated the power of machine learning over complicated tasks, but also provided the opportunity of artificial intelligence/computational intelligence to play a critical role in a wide range of applications.

This talk aims to discuss the recent research developments in integrated learning and control based on reinforcement learning (RL), one of the core foundations that AlphaGo/AlphaGo Zero was develop upon. Specifically, I will introduce a new RL and adaptive dynamic programing (ADP) framework for improved decision-making capability, and further explore their wide applications in cyber physical systems (CPS) across different domains. This framework integrates a hierarchical goal generator network to provide the system a more informative and detailed internal goal representation to guide its decision-making. Compared to the existing methods with a manual or “hand-crafted” reinforcement signal design, this framework can automatically and adaptively develop the internal goal representation over time. Under this framework, I will present numerous applications ranging from smart grid control to human-robot interaction to demonstrate its broader and far-reaching applications in CPS. As a multi-disciplinary research area, I will also discuss the future research challenges and opportunities in this field.


特邀讲者:Qingfu Zhang (张青富)

张青

Qingfu Zhang is a Professor at the Department of Computer Science, City University of Hong Kong. His main research interests include evolutionary computation, optimization, neural networks, data analysis, and their applications. He is currently leading the Metaheuristic Optimization Research (MOP) Group in City University of Hong Kong. Professor Zhang is an Associate Editor of the IEEE Transactions on Evolutionary Computation and the IEEE Transactions Cybernetics. MOEA/D, a multiobjective optimization algorithm developed in his group, won the Unconstrained Multiobjective Optimization Algorithm Competition at the Congress of Evolutionary Computation 2009, and was awarded the 2010 IEEE Transactions on Evolutionary Computation Outstanding Paper Award. He is on the list of the Thomson Reuters 2016 and 2017 highly cited researchers in computer science. He is a Fellow of IEEE. He is a Changjiang chair professor and was selected in 1000 talent program.

报告摘要:

进化计算是计算智能的重要分支,其早期发现常受自然进化或其它生物群体行为启发。进化计算已成为众多应用领域的重要优化方法,但其理论发展严重滞后。本报告将试图讨论进化优化方法与传统优化的关系,将主要探讨如下这几个问题:1. 遗传算法与梯度法之关系,2. 蚁群算法与梯度法之关系,3. Cma-es 与牛顿法之关系,4.多目标进化算法与传统分解方法之关系。对这些关系的深入研究也许将对进化计算的进一步发展起到积极推动作用。


特邀讲者:公茂果

公

公茂果,西安电子科技大学教授,博士生导师,计算智能研究中心主任,校学术委员会委员,陕西省重点科技创新团队负责人,国家重点研发计划项目首席。曾获"国家高层次人才特殊支持计划"中组部青年拔尖人才、国家优秀青年科学基金、霍英东青年教师奖、教育部新世纪优秀人才支持计划等。

公茂果教授主要研究方向为计算智能理论及其在数据与影像分析中的应用。主持国家重点研发计划、国家863计划、国家自然科学基金等十余项课题,发表SCI检索论文100余篇,被引用6000余次,入选中国高被引学者,授权国家发明专利20余项,获2013年国家自然科学奖二等奖和2016年教育部自然科学奖二等奖。担任IEEE IEEE Transactions on Evolutionary ComputationIEEE Transactions on Neural Networks and Learning Systems等期刊编委,IEEE计算智能学会Task Force on Collaborative Learning and Optimization主席,第十/十一届BIC-TA等学术会议主席,中国人工智能学会理事等。

报告摘要:面对大数据的诸多挑战,深度神经网络借助其深层结构,具备很强的复杂问题建模能力,在计算机视觉、人机对弈等很多应用中取得了突破性的进展。然而,深度神经网络在理论研究上仍然存在亟待解决的瓶颈难题。首先深度网络的结构设计困难,如网络层数、节点数目等都需要人工设定;同时,模型的表达参数对性能的影响显著,需要反复调参;而且,基于梯度的网络优化方法存在梯度弥散和陷入局部最优的缺点。本报告将介绍利用演化多目标优化解决上述难题的一些尝试,并汇报在深度神经网络解决空时影像变化检测关键难题上的一些最新进展。


热忱欢迎全国各地CCF会员、YOCSEF分论坛同仁,以及对本主题感兴趣的人士参加,期待您的光临!

参会方式:免费参加,食宿及交通费自理。

活动地址: 湖南省长沙市岳麓区麓山南路国家超算长沙中心

联系人:黄磊 17763630972 杨翠娟 18677381559  



CCF聚焦