The last decade has brought rapid advances in the performance of machine-learning systems: computers can now beat the best human players in Chess or Go, the seminal protein-folding problem has been solved computationally, and convolutional neural networks run on our smart phones. While this practical progress continues to happen and intelligent systems begin to appear at all places of society, it is especially important to understand how these systems work. Unfortunately, many ML systems are still comparably badly understood.
This lack of understanding has multiple bad effects. Unintended erratic behavior can sometimes not be excluded; costly and cumbersome manual tuning is often necessary; and engineering efforts can be directed at infeasible problems. Since modern ML stystems builds on techniques from multiple disciplines (statistics, optimization, numerical analysis, etc.), the interdisciplinary collaboration between expert is especially important to solve these challenges.
This workshop organized by the ELLIS Theory, Algorithms and Computations of Modern Learning Systems group is designed as a forum of exchange between some of Europe's leading researchers in the theory of ML. It aims to convey a comprehensive picture of ongoing researchers efforts to analyze ML systems and give orientation to researchers in the rapidly changing ML landscape.
The workshop agenda thus features talks from upcoming and established researches from diverse backgrounds (including Bayesian statistics, optimization, kernel methods, neural networks, etc.). In addition, there will be discussions and panels on the future of the ELLIS Theory, Algorithms and Computations of Modern Learning Systems group and on scientific topics of shared interest.
There will be ample time for brainstorming, networking and collaborating.
The 2022 ELLIS Theory Program Workshop is sponsored by ELISE, the European Network of AI Excellence Centres.
NOTE: The 2022 ELLIS Theory Program Workshop is an invitation-only event, and it is not possible to transfer this invitation to others. No fee is required.
The Workshop will take place from the 20th to the 22nd of June 2022. Activities will start on Monday the 20th afternoon and end, with a lunch, on Wednesday the 22nd.
The workshop will take place at Grand Hotel Arenzano, in Arenzano. The spaces provided by the Hotel follow the highest standards of safety precautions against Covid19, and will guarantee a safe and comfortable experience for all participants.
Arenzano is in the region of Liguria in the Italian Riviera (see here for a nice video).
The booking of single rooms at a reduced price at Grand Hotel Arenzano has been handled via a custom link, which has now expired. Please contact the organizers if you still need to book a room, so that we can advise on the best solution. Options for reimbursement depend on each participant's affiliation and have been therefore shared individually.
Within the H2020 ELISE project, the Theory Program has funds to cover reasonable travel costs of the workshop participants.
Rules for reimbursements depend on the status of your university/company towards the ELISE project.
Centre de Recherche INRIA - SIERRA project-team
Departement d'Informatique de l'Ecole Normale Superieure
PSL Research University
University of Tübingen
Max Planck Institute for Intelligent Systems
Università di Genova
Istituto Italiano di Tecnologia
Massachusetts Institute of Technology
14:00–14:20 | Welcome by Lorenzo Rosasco |
14:20–16:20 | Talks 1 |
16:20–16:45 | Break |
16:45–18:00 | Poster Session |
9:30–12:40 | Talks 2 |
12:40–15:30 | Lunch and time for brainstorming/networking |
15:30–16:10 | Keynote |
16:10–16:50 | Plenary discussion on ELLIS Theory group |
16:50–17:20 | Break |
17:20–18:00 | Keynote |
18:00–19:00 | Scientific Panel Discussion |
9:30–12:40 | Talks 3 |
12:40–13:00 | Closing by Philipp Hennig |
REGISTRATIONS ARE NOW CLOSED. For any inquiries please write to Lynn Anthonissen and Giulia Casu
ELLIS PhD & Postdoc Program Coordinator, ELISE Coordinator Tübingen
lynn (dot) anthonissen (at) uni-tuebingen (dot) de
Lab Manager at MaLGa Center, Coordinator of the ELLIS Genoa Unit
gcasu (dot) malga (at) gmail (dot) com