Name: Artificial Intelligence and the Uncertainty challenge in Fundamental Physics
Start: 2023-11-27T09:00:00+01:00
End: 2023-12-01T17:00:00+01:00
Location: No location set

Artificial Intelligence and the Uncertainty challenge in Fundamental Physics

from Monday, November 27, 2023 (9:00 AM) to Friday, December 1, 2023 (5:00 PM)

Monday, November 27, 2023

1:00 PM Welcome coffee : le Patio
Welcome coffee : le Patio
1:00 PM - 2:00 PM
2:00 PM Opening session, Uncertainty Quantification: Jussieu, Amphi 25
Jussieu, Amphi 25
2:00 PM - 6:00 PM
Contributions

2:00 PM Welcome - David Rousseau (IJCLab, CNRS/IN2P3, Université Paris-Saclay)

2:15 PM Uncertainty modeling in particle physics - Wouter Verkerke (Nikhef/UvA)

3:00 PM Uncertainty Quantification and Machine Learning in Industry: Current Practices and Challenges in Industrial Applications for Low-Carbon Electricity Production - Vincent CHABRIDON (EDF R&D)

3:30 PM An Introduction to Bayesian Optimization - Emmanuel Vazquez (L2S -- Paris-Saclay)

4:30 PM Uncertainty Quantification in Neural Networks: Methods and Considerations - Laurens Sluijterman (Radboud University)

4:50 PM Fair Universe Challenge Paris 2023 - Wahid Bhimji (Lawrence Berkeley National Laboratory)

5:20 PM Metrics for Uncertainty-Aware ML Methods - Sascha Diefenbacher (Lawrence Berkeley National Laboratory)
6:00 PM Welcome cocktail : le Patio
Welcome cocktail : le Patio
6:00 PM - 8:00 PM
Tuesday, November 28, 2023
9:00 AM Explainable AI: Amphi Durand, batiment Esclangon
Amphi Durand, batiment Esclangon
9:00 AM - 11:00 AM
Contributions

9:00 AM xAI in practice: current state of the art, limitations and perspectives - Julien Girard-Satabin (CEA)

9:45 AM Explainable AI for Interpretability of Deep Neural Networks : the High Energy Physics perspective - Mark Neubauer (University of Illinois at Urbana-Champaign)

10:35 AM Advancing Explainable AI: Testing and Enhancing Techniques Across Multidisciplinary Use-Cases - Simone Scardapane (Sapienza University of Rome, Roma, Italy)
11:00 AM Coffee break : espace 66-65
Coffee break : espace 66-65
11:00 AM - 11:30 AM
11:30 AM Simulation Based Inference: Amphi Durand, batiment Esclangon
Amphi Durand, batiment Esclangon
11:30 AM - 12:30 PM
Contributions

11:30 AM An introduction to simulation-based inference - Gilles Louppe (University of Liège)
12:30 PM Buffet at the top of Tour Zamansky, Jussieu
Buffet at the top of Tour Zamansky, Jussieu
12:30 PM - 2:00 PM
2:00 PM Simulation Based Inference: Amphi Astier, bâtiment Esclangon
Amphi Astier, bâtiment Esclangon
2:00 PM - 6:00 PM
Contributions

2:00 PM Simulation Based Inference: The Frequentist Perspective Abstract: - Harrison Prosper (Florida State University)

2:45 PM Simulation-Based Inference: Where Classical Statistics Meets Machine Learning - Mikael Kuusela (Carnegie Mellon University)

4:00 PM Swyft: Direct marginal inference for large simulation models - Christoph Weniger (University of Amsterdam)

4:30 PM Precision-Machine Learning for the Matrix Element Method - Nathan Huetsch (Heidelberg University)

5:00 PM Optimal dataset-wide inference in the presence of systematic uncertainties - Philipp Windischhofer (University of Chicago)
Wednesday, November 29, 2023
9:00 AM Frugal Unmixing, Reusing, and Sampling
9:00 AM - 11:00 AM
Contributions

9:00 AM Introduction at Institut Pascal

9:10 AM Data frugal machine learning approaches for unmixing problems in Physics - Jerome Bobin (CEA)

9:50 AM Publication and reuse of ML models in LHC analyses - Sabine Kraml (LPSC Grenoble, CNRS/IN2P3)

10:20 AM Machine Learning Assisted Sampling: Applications in Physics - Marylou Gabrié (Ecole Polytechnique)
11:00 AM Coffee break
Coffee break
11:00 AM - 11:30 AM
11:30 AM Benchmark, Datasets and Challenges
11:30 AM - 12:30 PM
Contributions

11:30 AM Exploring Data Challenges and Leveraging Codabench: A Practical Journey with unsupervised New Physics detection at 40 MHz - Ekaterina Govorkova (MIT)

12:00 PM Neuroscience ML challenges using CodaBench: Decoding multi-limb trajectories from two-photon calcium imaging - Yuan-Tang Chou (University of Washington, Seattle (US))
12:30 PM Buffet at Institut Pascal
Buffet at Institut Pascal
12:30 PM - 2:00 PM
2:00 PM Fair-Universe hackathon: Fair uni
Fair uni
2:00 PM - 6:10 PM
Contributions

2:00 PM Introduction - Wahid Bhimji (Lawrence Berkeley National Laboratory)

2:10 PM Overview of HiggsML Uncertainty Challenge - Ragansu Chakkappai (IJCLab-Orsay, CNRS/IN2P3, Université Paris-Saclay)

2:30 PM Walkthrough of CodaBench and submissions - Ihsan Ullah (ChaLearn)

3:00 PM Scoring and Baseline Systematic Aware method - Sascha Diefenbacher (Lawrence Berkeley National Laboratory)

3:50 PM Hands-on hackathon and discussion

4:30 PM Feedback from partipants

4:50 PM Wrap up and next steps - Wahid Bhimji (Lawrence Berkeley National Laboratory)
Thursday, November 30, 2023
9:00 AM Unfolding (de-biasing)
9:00 AM - 11:40 AM
Contributions

9:00 AM Unfolding in High Energy Physics - Vincent Alexander Croft (LIACS)

9:45 AM An unfolding method based on conditional Invertible Neural Networks (cINN) using iterative training - Mathias Josef Backes (Kirchhoff Institut für Physik)

10:15 AM Can contrastive learning de-bias my model? - Radi Radev (CERN)

11:15 AM Uncertainty Quantification and Anomaly Detection with Evidential Deep Learning - Mark Neubauer (University of Illinois at Urbana-Champaign)
11:40 AM Controlling uncertainties in generative models
11:40 AM - 12:25 PM
Contributions

11:40 AM Potential and challenges of highly dimensional generative models - Gregor Kasieczka (Universität Hamburg)
12:25 PM Buffet at Institut Pascal
Buffet at Institut Pascal
12:25 PM - 2:00 PM
2:00 PM Controlling uncertainties in generative models
2:00 PM - 6:00 PM
Contributions

2:00 PM Machine-learning and equations-informed tools for generation and augmentation of turbulent data. - Luca Biferale (U Roma Tor Vergata)

2:45 PM Generative modeling in genomics and a perspective on uncertainty quantification - Burak Yelmen (University of Tartu)

3:30 PM Uncertainty-aware diffusion models for LHC Event Generation - Sofia Palacios Schweitzer (ITP, Heidelberg University)

4:30 PM Data driven background estimation in HEP using generative adversarial networks - Mehmet Ozgur Sahin (Irfu, CEA Paris-Saclay)

5:00 PM Using an adversary trained on a control sample to control systematic errors - Gordon Watts (University of Washington)
7:00 PM Workshop Dinner at Brass&Co
Workshop Dinner at Brass&Co
7:00 PM - 9:00 PM
Friday, December 1, 2023
9:00 AM Architectures
9:00 AM - 12:00 PM
Contributions

9:00 AM Robustness to Uncertainties in Machine Learning Applications for HEP - tommaso dorigo (INFN Sezione di Padova)

9:45 AM Data Subsampling for Bayesian Neural Networks - Eiji Kawasaki (CEA)

10:15 AM Efficient Sampling from Bayesian Network Posteriors for Optimal Uncertainties - Sebastian Bieringer (Hamburg University, Institute for experimental physics)

11:15 AM Revisiting models and uncertainty with AI - Gael Varoquaux (INRIA-Saclay)
12:00 PM Buffet at Institut Pascal
Buffet at Institut Pascal
12:00 PM - 1:30 PM
1:30 PM Closing session: (starts at 1:30!)
(starts at 1:30!)
1:30 PM - 4:30 PM
Contributions

1:30 PM Highlights from the 1st CNRS AISSAI Thematic Quarter on Causality - Alessandro Leite (LISN, INRIA)

2:15 PM Fair Universe HiggsML Uncertainty Challenge: Lessons Learned and plans - Elham E Khoda (University of Washington)

2:45 PM Farewell - David Rousseau (IJCLab, CNRS/IN2P3, Université Paris-Saclay)

Choose timezone