SOCR News APS GDS 2024

Annual APS March Meeting

The annual American Physical Society's March Meetings are scientific research conferences convening over 13,000 physicists and students from around the globe to connect and collaborate across academia, industry, and major labs. The 2024 APS March meeting will celebrate the 125th anniversary of APS.

Overview

There are significant recent advances linking model-based statistical inference, quantum physics, computable data science, and model-free artificial intelligence (AI). Interlacing research from these areas has significant potential to lead to novel techniques for understanding and interpreting the physical world. Statistical physics methods model the behavior and interactions of large systems of objects and are applicable far beyond elementary particles. Computational data science algorithms facilitate the pre- and post-processing and analysis of heterogeneous types of information, including structured and unstructured, static and longitudinal, high- and low-dimensional, and complete and missing datasets. Model-free machine learning and AI techniques provide complementary approaches for exploratory, hypotheses-generating, data-mining, clustering, and predictive analytics based on few a priori assumptions. This session will bring a diverse pool of experts to discuss the theoretical challenges, empirical evidence, and potential opportunities of combining transdisciplinary methods to create meta ensemble techniques for holistic comprehensive understanding of complex systems.

2024 APS March GDS Invited Session

Invited Special Session Organizer & Chair

• Ivo Dinov, University of Michigan, SOCR, MIDAS.

Session Logistics

APS March 2024 Meeting

• Date/Time: Mar 7 2024 3:00PM CT
• Venue: Minneapolis Convention Center, Minneapolis, MN, USA, Room 205AB.
• Registration: Meeting Registration is required.
• Conference: March 2024 APS Annual Conference.
• Abstract Submission
• Session: Invited presentations - Session W56 Model-based Statistical Physics, Computable Data, and Model-Free Artificial Intelligence and W56 Session Summary
• APS March 2024 Invited Sessions

Program

Logistics

• Sponsoring Unit: APS Group on Data Science (GDS)
• Chair: Ivo Dinov, University of Michigan
• Venue: Minneapolis Convention Center, Room: 205AB, 1301 Second Ave S, Minneapolis, MN 55403
• Date/Time: Thursday, March 7, 2024; 3:00PM - 6:00PM

Time	Abstract	Invited Speaker
3:00PM - 3:30PM	W56.00001: On the use of physics in machine learning for imaging and quantifying complex processes	George Barbastathis
3:36PM - 4:06PM	W56.00002: Energy Frontier Exploration using Particle Physics and AI	Mark S Neubauer
4:12PM - 4:42PM	W56.00003: Data-driven medical image formation without a priori models	Michael Insana
4:48PM - 5:24PM	W56.00004: The Restricted Boltzmann Machine: from the statistical physics of disordered systems to a practical and interpretative generative machine learning	Aurélien Decelle

Abstracts

Abstract W56.00001: On the use of physics in machine learning for imaging and quantifying complex processes

• Time: 3:00 PM–3:36 PM
• Presenter: George Barbastathis (MIT)
• Co-Authors:
  • Qihang Zhang (Singapore-MIT Alliance for Research and Technology Centre; present address: Tsinghua University)
  • Richard D Braatz (Massachusetts Institute of Technology MIT)
  • Allan Myerson (Massachusetts Institute of Technology)
  • Charles Papageorgiou (Takeda Pharmaceuticals)
  • Wenlong Tang (Takeda Pharmaceuticals)
  • Yi Wei (Massachusetts Institute of Technology)
  • Neda Nazemifard (Takeda Pharmaceuticals)
  • Deborah Pereg (Massachusetts Institute of Technology)
  • Ajinkya Pandit (Massachusetts Institute of Technology)
  • Shashank Muddu (Massachusetts Institute of Technology)
  • Sandip Mondal (Sinagpore-MIT Alliance for Research and Technology Centre)
  • Daniel Roxby (Singapore-MIT Alliance for Research and Technology Centre)
  • Jongyoon Han (Massachusetts Institute of Technology MIT)

We discuss the use of machine learning kernels as regularizers in problems of quantitative imaging and estimation for complex processes. In such problems, the “image” is not a final goal; it is rather an intermediate step toward estimating parameter values.

Recently, we have investigated laser speckle as an encoder for dynamics of interacting particulates. For instance, we developed a real-time method, the Physics Enhanced Auto Correlation Estimator (Peace) [1] which explicitly maps the probability density function of particle sizes (also referred to as particle size distribution, PSD) to the intensity autocorrelation of the speckle. After recording the speckle and computing its autocorrelation, a machine learning-aided decoder returns the PSD. It is important to note that this is a far-field, or “non-imaging” method—the interpretation of speckle as an encoder obviates the need to image individual particles. Instead, the ensemble statistical properties are obtained directly from the autocorrelation.

In this talk, we will discuss more extensively the quantitative properties of dynamic Peace, i.e. when the particle size distribution itself is evolving due to chemical or mechanical interactions. We will also discuss some preliminary work on the application of quantitative speckle to two novel domains: scattering from biological cells and the phloem in plants. In both cases, the speckle is interpreted as an encoder of diffusion, transport and reactive processes. These may only be partially explained from first principles, whereas the constitutive relationships necessarily need to be derived from the data.

• References:
  • [1] Qihang Zhang, et al, Nature Comm. 14:1159, 2023.

• Acknowledgement: This research was funded by the National Research Foundation (NRF) of Singapore through the Intra-Create grant programme, grant no NRF2019-THE002-0006; and by Millennium Pharmaceuticals, Inc. (a subsidiary of Takeda Pharmaceuticals), grant No. D824/ MT15.

Abstract: W56.00002 : Energy Frontier Exploration using Particle Physics and AI

• Time: 3:36 PM–4:12 PM
• Presenter: Mark S Neubauer (University of Illinois at Urbana-Champaign)

Artificial Intelligence (AI) and machine learning (ML) methods have proven to be powerful tools for the exploration of physics at the energy frontier of particle physics. Their expanding role in fundamental physics is driven by the challenges of increasingly large and complex data from experiments and computationally expensive simulations required to model and interpret the data, in addition to the rapid development of more powerful AI/ML tools for science-driven data exploration and interpretation. Moreover, exploiting symmetries inherent in physics data have inspired physics-informed ML as a vibrant sub-field of computer science research. I In this talk, I will provide a brief overview of key applications of AI/ML to fundamental physics research at the energy frontier of particle physics and describe several future directions in areas including explainable AI, uncertainty quantification, anomaly detection and real-time AI systems that will significantly enhance the scientific capabilities and opportunities of future experiments. Finally, I will briefly touch on how we are entering a new era in the relationship between AI and science and how scientists will need to learn how to navigate in this new environment.

• Acknowledgements: We acknowledge support through the NSF cooperative agreement OAC-2117997 and the DOE Office of Science, Office of High Energy Physics, under Contract No. DE-SC0023365.

Abstract: W56.00003 : Data-driven medical image formation without a priori models

• Time: 4:12 PM–4:48 PM
• Presenter: Michael Insana
• Co-Authors: Will Newman (University of Illinois at Urbana-Champaign)

The neural networks learn material properties from sparse ultrasound measurements because every estimate of displacement made within a slowly deformed tissue contains information about the properties and boundary conditions throughout the entire contiguous medium. As an operator probes tissues and numerical model development begins, an entropy-based measure of data diversity indicates to the operator when sufficient information has been collected. Once a numerical model has converged, a constitutive model is applied to it to estimate parametric images for medical diagnosis. Unlike deep learning methods that require training data spanning any object that might be encountered, this method focuses on collecting a comprehensive data set for a given patient. We aim to discover what constitutes the comprehensive data set that informs machine learning for high-quality image formation.

Abstract: W56.00004 : The Restricted Boltzmann Machine: from the statistical physics of disordered systems to a practical and interpretative generative machine learning

• Time: 4:48 PM–5:24 PM
• Presenter: Aurélien Decelle (Universidad Complutense de Madrid)
• Co-Authors:
  • Beatriz Seoane (Univ Complutense)
  • Lorenzo Rosset (Laboratoire de physique de l'Ecole normale supérieure (LPENS))
  • Cyril Furtlehner (INRIA Paris Saclay)
  • Nicolas Bereux (LISN, Université Paris Saclay)
  • Giovanni Catania (Theoretical Physics department, Universidad Complutense de Madrid)
  • Elisabeth Agoritsas (University of Geneva)

In this talk I will present our recent work on the Restricted Boltzmann Machine (RBM). RBMs were introduced decades ago by Hinton as a variant of the Boltzmann Machine (BM), but with hidden variables and a characteristic bipartite architecture. RBMs, introduced at the time as a "product of experts", were successfully trained as a generative model using the so-called contrastive-divergence method and, despite their shallow and simple architecture, were able to generate convincing new samples for complex real-world datasets. They later became popular as building blocks for pre-trained deep neural networks before the advent of more sophisticated methods.

The increasing interest of physicists and statistical physicists in RBMs in recent years is driven to several factors. First, RBMs can be seen as a generalization of BM that can be used to study interesting emerging phenomena, such as the phase diagram of the learned machine, how the learned free energy landscape is related to the properties of the dataset, or how the features of the dataset are encoded during the learning dynamics. Secondly, its practicality and simplicity make it an accessible model for physicists, providing a more understandable alternative to large, opaque neural networks.

I will present our understanding of the phase diagram and the learning dynamics of this model at both analytical and numerical levels. I will then show how we can construct equivalences between RBMs and generalized BMs where the weights of the RBM can be mapped into effective K-body interactions so that we are able to infer interacting components for a given dataset.

• Acknowledgements: Comunidad de Madrid and the Complutense University of Madrid through the Atracción de Talento programs (Refs. 2019-T1/TIC-13298)The Banco Santander and the UCM (grant PR44/21-29937)Ministerio de Economía y Competitividad, Agencia Estatal de Investigación and Fondo Europeo de Desarrollo Regional (Ref. PID2021-125506NA-I00).

Resources

• ... coming up later ...

---

---

---

Translate this page:

(default)	Deutsch	Español	Français	Italiano	Português	日本語	България	الامارات العربية المتحدة	Suomi	इस भाषा में	Norge
한국어	中文	繁体中文	Русский	Nederlands	Ελληνικά	Hrvatska	Česká republika	Danmark	Polska	România	Sverige