BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:AI Center Seminar - AI Fundamentals series - Prof. Emir Demirović
DTSTART:20251028T101500
DTEND:20251028T111500
DTSTAMP:20260502T074756Z
UID:16309978c1ba8a12d22c9583a1ec91238975d45b9398bef182e989b1
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Emir Demirović\nThe talk is organized by the EPFL AI 
 Center as part of the AI fundamentals seminar series.\n\nHosting professo
 r: Prof. Clément Pit-Claudel\n\nTitle\nTransparent AI by Design: Search A
 lgorithms for Supervised Learning\, Control Policies\, and Combinatorial C
 ertification\n\nAbstract\nAI methods—such as those used in supervised le
 arning\, controller synthesis\, and combinatorial optimisation—have demo
 nstrated immense value across many domains. However\, their practical adop
 tion is hindered by reliability concerns\, particularly when these systems
  are designed as black boxes. Two key challenges arise for black-box AI: (
 1) lack of performance guarantees—when AI fails\, it is unclear whether 
 the task is infeasible or the underlying algorithm is simply inadequate\; 
 and (2) lack of confidence—results may be difficult to interpret or trus
 t. While post-hoc interpretability techniques offer partial remedies\, we 
 advocate for a different paradigm: building AI systems that are transparen
 t by design. Rather than explaining opaque decisions after the fact\, we s
 ynthesise outputs that are intrinsically understandable and verifiable. Th
 is shifts the focus from doubting AI to questioning whether we are solving
  the right problem. We apply this approach across three distinct domains: 
 supervised learning\, controller synthesis\, and infeasibility certificati
 on for combinatorial optimisation problems. Although these tasks involve e
 xponentially large search spaces\, recent advances demonstrate that design
 ing for transparency is increasingly practical—often without sacrificing
  performance—making it a compelling alternative to opaque AI systems. A 
 variant of this talk will be presented at ECAI'25 as an invited talk in th
 e Frontiers in AI series.\n\nBio\nDr. Emir Demirović is an assistant prof
 essor of computer science at TU Delft (Netherlands)\, where he leads the C
 onstraint Solving ("ConSol") research group and directs the Explainable AI
  in Transportation ("XAIT") lab. He has been recognised with the Early Car
 eer Researcher Award 2025 from the Association for Constraint Programming 
 and is an ELLIS Scholar. His research focuses on exploiting structural pro
 perties of NP-hard problems to design algorithms that are both theoretical
 ly complete and efficient in practice\, with a particular emphasis on cons
 traint programming and dynamic programming techniques. His techniques have
  advanced state-of-the-art solvers in MaxSAT and constraint programming (a
 chieving high rankings in competitions such as the MaxSAT Evaluation and t
 he MiniZinc Challenge)\, his optimal decision tree methods (machine learni
 ng) are among the fastest\, and his recent approach to certifying outputs 
 of constraint programming solvers on large-scale problems has been highlig
 hted by Donald Knuth in The Art of Computer Programming (Vol. 4\, Fasc. 7)
  as a promising advancement.
LOCATION:ELE 117 https://plan.epfl.ch/?room==ELE%20117 https://epfl.zoom.u
 s/j/67720676571?pwd=2cvgMoHxhzvdKS4PUCZoZCCkYueSvg.1
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR