1st workshop on AI and Mathematics - connecting the mathematics clusters

 

The first AIM workshop is organised by the Dutch Mathematics Clusters and its goal is to highlight the role of mathematics as a key enabling technology within the emerging field of scientific machine learning, and bring together researchers across mathematics.

Artificial Intelligence (AI) will have a growing impact on all sciences and business sectors, our private lives, and society as a whole. It is pre-eminently a multidisciplinary technology that connects scientists from a wide variety of research areas, from behavioural science and ethics to mathematics and computer science. Without downplaying the importance of its interdisciplinary nature, it is apparent that mathematics can and should play an active role. As Robbert Dijkgraaf observed in NRC in May 2019: ''Artificial intelligence is in its adolescent phase, characterised by trial and error, self-aggrandisement, credulity and lack of systematic understanding''. Mathematics can contribute to this much-needed systematic understanding of AI and at the same time lay the ground work for further improvements.

In the two-day program we showcase recent research on the interface between mathematics and AI. There will be plenty of time for informal discussions, as well as strategic sessions.

  Registration
The workshop is free of charge after registration, please register here.

If you have any questions, please contact us at: aim@nwo.nl

Accommodation
Hotel reservation can be booked directly at Hotel Casa for a corporate rate of 119 euro including breakfast. Please use the promotion code AI&Mathematics.

Program (flyer)

Thursday 9 June

8:00 – 9:00        Registration

9:00 – 9:10        Opening Christoph Brune

9:10 – 10:40      DIAMANT chair: Steven Kelk

              Marten van Dijk – CWI (30 min)

              On the Convergence of Stochastic Gradient Descent

              Ronald de Wolf – CWI/University of Amsterdam (30 min)

              Mathematical Aspects of Quantum Machine Learning

              Antonios Antoniadis – University of Twente (30 min)

              Learning-Augmented Algorithms for Dynamic Power Management

10:40 – 10:55    Coffee break

10:55 – 12:25    STAR chair: Evgeny Verbitskiy

              Peter Grünwald – CWI/Leiden University (30 min)

              The Unreasonable Effectiveness of Stochastic Mathematics in Artificial Intelligence

              Tim van Erven – University of Amsterdam (30 min)

              Statistics and Machine Learning: Towards a Closer Integration

              Bert Zwart – CWI/Eindhoven University of Technology (30 min)

             Machine Learning and applied probability

12:25 – 13:45    Lunch break

13:45 – 15:15    GQT chair: Christoph Brune

              Pepijn Roos Hoefgeest – Vrije Universiteit Amsterdam (45 min)

              An introduction to persistent homology

              Bram Mesland – Leiden University (45 min)

              Quantum geometry in noisy data

15:15 – 15:30    Coffee break

15:30 – 17:00    NDNS+ chair: Tristan van Leeuwen

              Gitta Kutyniok – LMU Munich (45 min)

              Scientific Computing meets Artificial Intelligence

              Remco Duits – Eindhoven University of Technology (45 min)

              PDE-G-CNNs: PDE-based roto-translation equivariant convolutional neural networks and applications

17:00 – 18:30     Poster session and drinks

18:30 – 21:00     Workshop dinner at Maslow, Carolina MacGillavrylaan 3198, 1098 XK Amsterdam

 

Friday 10 June

8:00 – 9:00      Arrival and registration

9:00 – 10:00     DIAMANT chair: Mathias Staudigl

              Etienne de Klerk – Tilburg University (30 min)

              Analyzing the worst-case behavior of popular optimization algorithms used in machine learning

              Ilker Birbil – University of Amsterdam (30 min)

              A Scalable Rule Generation Framework for Learning

10:00 – 11:00    STAR chair: Tim van Erven

              Jaron Sanders – Eindhoven University of Technology (20 min)

              Analyzing dropout training in neural networks from a stochastic perspective

              Rianne de Heide – Vrije Universiteit Amsterdam (20 min)

              Stochastic bandits

              Arnoud den Boer – University of Amsterdam (20 min)

              Cartel formation by data-driven price algorithms

11:00 – 11:15     Coffee break

11:15 – 12:15    GQT chair: Bram Mesland

              Ieke Moerdijk  – Utrecht University (60 min)

              Shuffles of Trees

12:15 – 13:30     Lunch break

13:30 – 14:30     NDNS+ chair: Tristan van Leeuwen

              Felix Lucka – CWI (20 min)

              Deep Learning in Computational Imaging

              Palina Salanevich – Utrecht University (20 min)

              Online non-negative matrix factorization as a tool in data processing

             Marcello Carioni – University of Twente (20 min)

              Learning adversarial regularizers for inverse problems

14:30 – 17:00     Strategic session chair: Wil Schilders

              Inald Langedijk – AiNed (15 min)

              Mengwu Guo – 4TU-AMI (15 min)

             Strategic Research Initivatives proposal Discussion

              Group discussions

17:00 – 18:00     Farewell drinks

 The program booklet with abstracts is available here.

Organizing team

  • Christoph Brune (University of Twente), chair
  • Leo van Iersel (TU Delft, co-chair), Mathias Staudigl (UM) and Steven Kelk (UM), representatives of DIAMANT
  • Wil Schilders (TU/e) and Tristan van Leeuwen (CWI), representatives NDNS+
  • Evgeny Verbitskiy (UL), representative STAR
  • Bram Mesland (UL), representative GQT
  • Olivia Muthsam (NWO)
  • Nada Mitrovic (CWI)