ERAI专题


第30届中国过程控制会议

 ERAI专题研讨会

地点:昆明世纪金源大饭店

主题:4届证据推理与人工智能专题会议 (ERAI2019)

主席: Jian-Bo Yang, Alliance Manchester Business School, The University of Manchester, UK.

嘉宾:

Dong-Ling Xu, Alliance Manchester Business School, The University of Manchester, UK.

胡昌华, 火箭军工程大学教授, 长江学者、杰青.

  , 北京华胜天成科技股份有限公司CTO.

吴建德, 昆明理工大学教授, 民航学院院长.

 

召集人:  斌,昆明理工大学教授,信息工程与自动化学院副院长

主要内容:会议包括上午主会场4个报告,下午3个分会场共18个报告。报告内容主要围绕不确定信息融合、证据推理、置信规则库、智能决策、智能检测与优化等主题开展,重点关注以上技术在智慧医疗与智能辅助诊断、工程系统安全性评估、大数据分析与网络安全、图像处理、工业软测量等领域的应用。

 

 

报告题目1: Hybrid Data-Driven and Knowledge-Based Fraud Detection via Evidential Reasoning

报告人:Dong-Ling Xu and Jian-Bo Yang

  要:Machine learning models have gained an increasing amount of attention in learning complex patterns. However, such learning capacity of complex pattern has come at the cost of model interpretability. In this presentation, we discuss a transparent intelligent system which can help make decisions by acquiring evidence and evaluating the reliability of evidence from training data. The Evidential Reasoning (ER) rule is used to combine multiple pieces of evidence gathered for a hierarchy of multiple indicators. Several strategies for constructing indicator hierarchy are explored by means of making appropriate use of both expert knowledge and data in statistically meaningful manner. Each piece of evidence is weighted and combined conjunctively with the weights optimised by using a maximum likelihood evidential reasoning (MAKER) framework for extended probabilistic inference. Supported by a multinational law firm and UK government, the system is applied to a real-world dataset on insurance claims. Results from the application are presented, which reveal that the proposed method preserves the interpretability and usability of expert detection systems, and can be used to predict changing patterns in fraud practices by tracking the trend of the weights of experience-based indicators. The findings show that the proposed method outperforms a number of widely used machine learning models.

个人简介:

Dr Dong-Ling Xu is chair professor of Decision Systems at Alliance Manchester Business School, The University of Manchester, UK. Over the past 30 years, she has been conducting research in data analysis, statistical inference, machine learning and decision support systems under uncertainty; system and process modelling using techniques from operational research, artificial intelligence, and management systems; statistical fault detection, system development and their applications in a wide range of areas. She has developed several interactive web-based decision support systems and co-developed a Windows-based software package called Intelligent Decision System (IDS) via Evidential Reasoning for general purpose multiple criteria decision analysis, and several statistical fault detection systems. Those systems are used in a wide range of decision making and risk assessment activities by organizations such as General Motors, Tesco, NHS, Ford, Shell, BP and CNOOC in areas such as healthcare and finance; modelling and analysis of system safety and security; and organisational self-assessment in quality management. The Evidential Reasoning approach and the IDS software is used by practitioners and researchers from over 50 countries. She has published over 100 peer reviewed journal papers, book chapters and books.

 

Dr Jian-Bo Yang is Chair Professor of Decision and Systems Sciences and Director of the Decision and Cognitive Sciences Research Centre at Alliance Manchester Business School of The University of Manchester, UK. He is also Changjian Chair Professor specially appointed by China Ministry of Education. Over the last three decades, he has been conducting research in a range of areas, including evidential reasoning theory; multiple criteria decision analysis under uncertainty; multiple objective optimization; decision support systems; hybrid quantitative and qualitative decision modelling using techniques from operational research, artificial intelligence and systems engineering; complex system modelling, simulation and control for engineering and management systems. His current applied research is in the areas of financial risk and strategic decision modelling and analysis; hybrid data-driven and knowledge based modelling and decision support for healthcare; new product development; aggregated production management; system maintenance management; risk and security modelling and analysis; performance analysis and improvement; the integrated evaluation of products, systems, projects, policies, etc. He has been awarded research grants at the total value of over £4m by the public funding bodies of the UK, EU, China and USA, such as the UK Engineering and Physical Science Research Council (EPSRC), European Commission (EC), the UK Department of Environment, Food and Rural Affairs (DEFRA), Natural Science Foundation of China (NSFC), China Ministry of Science and Technology, Hong Kong Research Grant Council (HKRGC) and the US Air Force Office of Scientific Research (AFOSR), as well as industry such as General Motors Company and Tesco PLC. He has published 4 books, over 200 journal papers and book chapters, and a similar number of conference papers, with high citations in Web of Science and Google Scholar, and developed several software packages including the Windows-based intelligent decision system (IDS) via evidential reasoning, which has been used by thousands of academics, researchers, practitioners and students over 50 countries.

 

 

报告题目2: 基于半定量信息的动态复杂系统建模理论及其故障预测

报告人:胡昌华

摘要:随着科学技术的进步,动态复杂系统的安全性迫切需要提高。故障预测同时具有提高系统安全性和节约维护成本的双重功效,其需要解决的关键问题是建立故障的在线预测模型。本报告主要介绍基于半定量信息的复杂系统建模理论,包括置信规则库与证据推理方法;分析现有故障预测方法存在的问题;提出动态ER算法以及BRB模型的参数优化和结构在线辨识方法,并展示上述理论成果在各领域中的应用。

个人简介:

火箭军工程大学教授,博士生导师,导航制导与控制国家重点学科带头人、导弹测试与控制技术国家虚拟仿真中心主任,长江学者特聘教授,国家杰出青年科学基金获得者,国家教学名师,军队领军人才,教育部自动化类教学指导委员会委员,中国自动化学会常务理事、技术过程故障诊断与安全性委员会副主任,军委科技委精确制导专家组专家,火箭军导弹专家,陕西省自动化学会副理事长。出版专著7部、教材1部,入选国家规划教材1部,授权国家发明专利8项,发表论文300余篇,SCI收录62篇,SCI他引1268次,获国家科技进步二等奖1项,军队(省)部级科技进步一、二等奖9项,当选全国优秀科技工作者,新世纪百千万人才国家级人选、国家有突出贡献的中青年专家,中国科协“求是”杰出青年实用工程奖,中国科协八大代表,享受政府特殊津贴专家,军队杰出专业技术人才奖,军队院校育才奖金奖,全军爱军精武标兵,火箭军十大砺剑尖兵。

 

 

报告题目3: 基于行为大数据的金融风险管控

报告人: 

摘要:金融风险管控是金融机构的核心竞争力,华胜天成在对金融客户的业务需求基础上,进行了大量的实践。本报告介绍如何通过对客户行为数据的深度学习,建立基于证据的风控与决策模型,进行风险分析,确保金融机构在合规、授信、反欺诈、反洗钱等业务处理过程的事前、事中、事后风险防范的有效性,从而体现金融大数据处理的核心价值。

个人简介:

北京华胜天成科技股份有限公司CTO,金融行业应用软件系统集成及应用软件开发近30年的从业经验,对国内金融行业信息化发展有深入的理解。

 

 

报告题目4: 高原山地长输管道服役行为与结构安全研究进展及展望

报告人:吴建德

摘要:本报告介绍高原山地长距离输送管道机理建模及安全监控关键技术的研究工作;揭示高原山地长输管道服役行为及结构安全保障技术;阐述数据驱动的高原山地长输管道核心设备状态监控及健康监测关键技术;分析高原山地长输管道服役行为与结构安全下一步发展重点和行动方向。

个人简介:

昆明理工大学民航学院院长,教授,博士生导师,云南省万人计划产业技术领军人才,云南省中青年学术和技术带头人,云南省首批专家基层科研工作站设站专家,中国有色金属工业智能系统与先进控制工程技术研究中心主任,云南省矿物管道输送工程技术研究中心主任,云南省高校复杂系统智能检测与控制科技创新团队首席教授。

 

    嘉宾: