报告名称：Trustworthy Systems: Machine learning and more formal challenges

报告时间：7月14日10:00

报告地点：滴水湖国际软件学院408室

报告摘要:

A key focus of his presentation will be a comprehensive study involving a cardiac arrest hotline, where his team undertook a thorough assessment of trustworthiness and ethical compliance. This evaluation highlights the critical need for trustworthy and ethical AI applications in high-stakes healthcare settings. In addition, Dr. Düdder will share insights from ongoing research that explores the complexities of trustworthy computation, which is essential for developing AI systems that users can depend on. He will discuss various challenges associated with integrating trustworthy computation into practical application contexts, shedding light on the obstacles that need to be overcome to ensure that these AI systems not only perform effectively but also uphold ethical standards and foster trust among users. Through this presentation, attendees can expect to gain a deeper understanding of the intersection between technology, ethics, and safety-critical applications.

报告人简介:

Dr. Boris Düdder is an associate professor of software engineering and formal methods at the Department of Computer Science (DIKU) at the University of Copenhagen (UCPH). He is the Deputy Head of Department for Teaching at DIKU, the Director of the Trustworthy AI Lab at UCPH, and a member of the working group on the General-Purpose AI Code of Practice at DG CNECT, the European Commission. He is widely recognised as a leading authority in the field of software engineering and formal methods. Dr. Düdder leads the research group on Software Engineering & Formal Methods at DIKU. Dr. Düdder’s primary research interests lie in the areas of formal methods and programming languages in software engineering of trustworthy distributed systems, with a focus on automated program generation for adaptive systems with high-reliability guarantees. His work also involves the computational foundations of reliable and secure Big Data ecosystems. Dr. Düdder’s long-term research vision is to develop dependable, adaptive, and software-defined technical systems based on program synthesis for manufacturing, healthcare, and logistics.