Learning to Discover

Session

Representation Learning workshop

REP

19 avr. 2022, 09:15

Institut Pascal

Présidents de session

Representation Learning workshop: Tues Morning

• David Rousseau (IJCLab, Orsay, France)

Representation Learning workshop: Tues afternoon

• Jean-Roch Vlimant (California Institute of Technology (US))

Representation Learning workshop: Wed morning

• Peter Battaglia (DeepMind)

Representation Learning workshop: Wed afternoon

• Andreas Salzburger (CERN)

45. Introduction from the organisers

David Rousseau (IJCLab, Orsay, France)

19/04/2022 09:15

46. Welcome from Institut Pascal director

Denis ULLMO (Institut Pascal)

19/04/2022 09:30

47. Introduction and review: machine learning for charged particle tracking

Jan Stark ( Laboratoire des 2 Infinis - Toulouse, CNRS / Univ. Paul Sabatier (FR))

19/04/2022 09:45

48. Geometric Deep Learning, Graph Neural Networks, and Neural Diffusion Equations

Michael Bronstein (Imperial College, London)

19/04/2022 11:15

Symmetry as an organising principle has played a pivotal role in Klein's Erlangen Programme unifying various types of geometry, in modern physics theory unifying different types of interactions. In machine learning, symmetry underlies Geometric Deep Learning, a group-theoretical framework for a principled design of geometric inductive biases by exploiting symmetries arising from the structure...

49. Overview of Machine Learning for Calorimeter and Particle Flow

Joosep Pata (NICPB, Tallinn)

19/04/2022 14:00

The reconstruction of particle signals relies on local reconstrution, which involves clustering of granular hits within detector subsystems, followed by global reconstruction, combining signals across detector subsystems for a high-level particle representation of the event. Calorimeter clustering is a local reconstruction method that aims to segment calorimeter hits according to their...

50. Learning general purpose physical simulators

Alvaro Sanchez (Deepmind)

19/04/2022 15:00

Simulations are central to modeling complex physical systems in many disciplines across science and engineering. However, high-dimensional scientific simulations can be very expensive to run, and require specialized solvers. In this talk, we will review some of our recent work on a general purpose framework for learning grid-based, particle-based, and mesh-based simulations using convolutional...

51. Break-out sessions

19/04/2022 16:30

52. Enabling Empirically and Theoretically Sound Algorithmic Alignment

Petar Velickovic (Deepmind / University of Cambridge)

20/04/2022 09:00

Neural networks that are able to reliably execute algorithmic computation may hold transformative potential to both machine learning and theoretical computer science. On one hand, they could enable the kind of extrapolative generalisation scarcely seen with deep learning models. On another, they may allow for running classical algorithms on inputs previously considered inaccessible to...

53. Breakout sessions

20/04/2022 11:00

54. Rediscovering orbital mechanics with Machine Learning

Pablo Lemos (University of Sussex)

20/04/2022 14:00

We present an approach for using machine learning to automatically discover the governing equations and hidden properties of real physical systems from observations. We train a "graph neural network" to simulate the dynamics of our solar system's Sun, planets, and large moons from 30 years of trajectory data. We then use symbolic regression to discover an analytical expression for the force...

96. Break-out session

20/04/2022 16:00

55. Conclusions from breakout sessions

20/04/2022 17:00