ATLAS Collaboration’s Post

Machine learning and AI - not just for generating odes to your dog. It can also be used to further our understanding of the universe. Brilliant work by the ATLAS Collaboration in gaining even better insights from CERN’s massive ATLAS data sets #AIisMoreThanGenAI

Recent work on physics-informed data clustering is published in the Journal of Computational Physics! In this work, we inject physical knowledge into the clustering procedure by using distances scaled by dynamical system Jacobians. We show how this biases the cluster distribution towards dynamically sensitive regions — a useful property for cluster-conditioned feature extraction and modeling. This work was done with Venkat Raman (University of Michigan). Read more here: https://lnkd.in/d8kqMMfM 

The September 2024 issue of ASME Journal of Computational and Nonlinear Dynamics includes research papers, and a technical brief, on neural ring networks, defect correction methods, and more: 🔹 Additional Natural Frequency of the Beam Carrying a Spring-Mass System: Lost and Found 🔹 Full-Dimensional Proportional-Derivative Control Technique for Turing Pattern and Bifurcation of Delayed Reaction-Diffusion Bidirectional Ring Neural Networks 🔹 Haar Wavelet Approach for the Mathematical Model on Hepatitis B Virus 🔹 Nonlinear Static and Dynamic Responses of a Floating Rod Pendulum 🔹 A Posteriori Error Analysis of Defect Correction Method for Singular Perturbation Problems With Discontinuous Coefficient and Point Source 🔹 An Improved Wiener Path Integral Approach for Stochastic Response Estimation of Nonlinear Systems Under Non-White Excitation 🔹 Harmonic Response of a Highly Flexible Thin Long Cantilever Beam: A Semi-Analytical Approach in Time-Domain With ANCF Modeling and Experimental Validation ASME Journal Program #ASMEJCND #nonlineardynamics https://lnkd.in/eXNxwCx4

Excited to share that a part of my bachelor thesis has been published in the Journal of Chemical Physics! In our paper, we explore the inverse design of crystals and quasicrystals in a non-additive binary mixture of hard disks. By using an evolutionary strategy (CMA-ES) to solve a black box optimization problem, we effectively identify state points where self-assembly occurs. A big thanks to Edwin A. Bedolla-Montiel, Alberto Pérez de Alba Ortíz, and Marjolein Dijkstra for their guidance throughout my bachelor thesis and the publication process. https://lnkd.in/eRHAZiik

Excellent article discussing a mathematical framework to understand emergence in complex systems. The framework uses computational mechanics to identify hierarchical structures within systems, allowing for the prediction of macro-level behavior without needing to consider micro-level information. The article also explores the implications of this understanding for concepts like free will and the structure of the universe. #ComplexityScience #SystemsScience #Emergence https://lnkd.in/gCnChHVA

For friends of "spooky action at a distance": Non-locality is important, both for the study of the foundations of quantum mechanics and for modern applications like device-independent quantum information processing. But which states exactly are non-local? Surprisingly, it still remains a challenge to answer this question for arbitrary states. With my PhD student Nick von Selzam, we are happy to present a new method to address this challenge. Our numerical technique explicitly constructs local hidden-variable models for arbitrary local quantum many-body states, borrowing tools from machine learning. Already now the approach leads to new insights, like the actual critical noise (visibility) at which a Werner state and other popular states become non-local. Since the method is not restricted to two parties, we also study the non-locality of quantum many-body states in correlated spin systems. Read more about it in our preprint: https://lnkd.in/eVCJn7XM Also, check out the github if you have states you want to test: https://lnkd.in/eRMB_pWp #LocalHiddenVariables #QuantumMechanics #BellInequalities Max Planck Institute for the Science of Light FAU Profile Center Light.Matter.QuantumTechnologies

Today I gladly share with my network that our paper entitled "A data-driven turbulence modeling for the Reynolds stress tensor transport equation" has been published at the International Journal for Numerical Methods in Fluids! It is comprised by a significant part of my master's thesis. I co-authored it with Matheus Altomare, MSc., Bernardo Brener and my thesis advisor Roney Thompson. In this work we have introduced a modified transport equation for the Reynolds stress that is driven by a source term predicted by neural networks. The transport equation was coupled with the momentum balance and the SIMPLE algorithm for pressure, forming a full data-driven Reynolds stress model, which was used to correct RANS simulations. DNS simulations for the square-duct flow were used to train the newtork and validate the results. You can access it at https://lnkd.in/dJaB76Ny and it is fully available for free at https://lnkd.in/dZTyKt8H You can also access the model's implementation as an OpenFOAM turbulence model at this repository on Github https://lnkd.in/dpb9UqRm

Perspective paper on simulations of water with machine learning potentials (Omranpour, Montero De Hijes, Behler & Dellago) https://lnkd.in/d3U5iJjY #MachineLearning #SimulationScience #WaterResearch #ComputationalChemistry #MolecularDynamics #MLPotentials #Chemistry #AIinScience #PhysicalChemistry #MaterialScience

Trinity project complete ! I've told you about our "Trinity" project: three tutorial papers about simulation, optimization, deep-learning based inverse design... well all the papers have been published and all of them are open access !! In collaboration with Peter Wiecha LAAS-CNRS and Olivier Teytaud Meta, and with Pauline Bennet and Denis Langevin from Institut Pascal. Paper #1 PyMoosh (simulation/database construction) https://lnkd.in/exX2fp5G Paper #2 Optimization in photonics (but if you're interested in optimization globally, probably worth reading anyway, we really give tips). https://lnkd.in/eyzuDSTP Paper #3 Inverse design & deep learning https://lnkd.in/ec6gfb5W I'm really proud. So much work also...

According to this calculation https://lnkd.in/gMZuXCBb, after all these years, the Large Hadron Collider experiments have explored less than 2% of event types (ATLAS Collaboration, Phys.org). Need Machine Learning 🔔