Quantifying Uncertainty in Machine Learning Interatomic Potentials for Accurate Molecular Dynamics Simulations
il y a 5 jours
**Quantifying Uncertainty in Machine Learning Interatomic Potentials for Accurate Molecular Dynamics Simulations**:
- Réf **ABG-128311**
- Sujet de Thèse
- 03/02/2025
- Contrat doctoral
- IFP Energies nouvelles
- Lieu de travail- Rueil-Malmaison - Ile-de-France - France
- Intitulé du sujet- Quantifying Uncertainty in Machine Learning Interatomic Potentials for Accurate Molecular Dynamics Simulations
- Champs scientifiques- Mathématiques
- Mots clés- machine learning, uncertainty quantification, molecular dynamics
**Description du sujet**:
In the pursuit of energy transition, optimizing catalytic processes is essential for the efficient and sustainable conversion of raw materials into biofuels. At the heart of this optimization lies a deep understanding of atomic-level reaction mechanisms. Molecular Dynamics (MD) simulations, coupled with quantum mechanical calculations (DFT), provide powerful insights into these mechanisms. Yet, the high computational cost of DFT presents a significant challenge.
To address this limitation, Machine Learning (ML) approaches have emerged as a promising solution, accelerating simulations by replicating DFT results at a fraction of the computational cost using Machine Learning Interatomic Potentials (MLIP). Despite their advantages, MLIP models are prone to accumulating errors during simulations, which can undermine the reliability of the results. This PhD aims to develop a robust uncertainty quantification framework to control and estimate errors in MLIP-based Molecular Dynamics simulations, enhancing the reliability of reaction rate predictions.
- Error Control Framework: Design an uncertainty quantification method that triggers DFT calculations when MLIP errors exceed a predefined threshold.
- Error Propagation in Kinetics: Develop techniques to propagate uncertainties through MD simulations, enabling accurate estimation of reaction rate constants.
- Extension to Nonlinear Models: Adapt the uncertainty quantification framework for nonlinear MLIP models, such as graph neural networks, to broaden its applicability to complex interatomic potentials.
The proposed methods will be validated using the dehydration of isobutanol catalyzed by acidic zeolites—a reaction of significant interest in biomass conversion due to its complexity and industrial relevance.
**Prise de fonction**:
- 03/11/2025
**Nature du financement**:
- Contrat doctoral
**Précisions sur le financement**:
**Présentation établissement et labo d'accueil**:
- IFP Energies nouvelles
IFP Energies nouvelles is a French public-sector research, innovation and training center. Its mission is to develop efficient, economical, clean and sustainable technologies in the fields of energy, transport and the environment. For more information, see our WEB site.
IFPEN offers a stimulating research environment, with access to first in class laboratory infrastructures and computing facilities. IFPEN offers competitive salary and benefits packages. All PhD students have access to dedicated seminars and training sessions.
**Site web**:
- **Academic requirements** University Master or Engineering School degree in applied mathematics, statistics/probability, machine learning
- **Language requirements** English level B2 (CEFR), willingness to learn French
- **Other requirements** solid knowledge in programming (Python), curiosity and critical thinking skills
-
Rueil-Malmaison, Île-de-France Axens Temps pleinLa Project Operations Business Division (POP) recherche une/unInternship : Reference Simulation for Fluidized Catalytic Cracking Process (F/M)Localisation : Rueil-MalmaisonAxens Presentation :Axens group provides a complete range of solutions for the conversion of oil and biomass to cleaner fuels, the production and purification of major petrochemical...
-
Rueil-Malmaison, France IFPEN Temps plein**Context** IFP Energies nouvelles (IFPEN) is a major player in research and training in the fields of energy, transport, and the environment. From research to industry, technological innovation is at the heart of all its activities. As part of the public-interest mission entrusted to it by public authorities, IFPEN is committed to providing solutions to...
-
Postdoctoral Position
il y a 1 semaine
Rueil-Malmaison, France IFPEN Temps plein**Contexte**: To comply with the Paris Agreement and achieve carbon neutrality by 2050, France aims to reduce its industrial sector's greenhouse gas emissions by 81% from 2015 levels by 2050. Various decarbonization strategies are available, including the adoption of new technologies like alternative fuels (e.g., hydrogen, ammonia) and CO2 capture, as well...
-
AI Industry Consulting Analyst
il y a 7 jours
Rueil-Malmaison, Île-de-France Schneider Electric Temps pleinSchneider Electric's purpose is to empower all to make the most of our energy and resources, bridging progress and sustainability for all. At Schneider, we call this Life Is On.We are driving digital transformation by integrating the most advanced energy and automation technologies. We connect products, control platforms, software and services throughout...
-
AI Industry Consulting Analyst
il y a 3 heures
Rueil-Malmaison, Île-de-France Schneider Electric Temps pleinBasic Qualifications:Masters or Ph.D. (Computer Science, Statistics, Engineering, Physics, Mathematics, Economics)Minimum 1 years of Data Science experienceMinimum 1 years of work experience in relevant domains (Energy, Oil & Gas, Chemicals, Natural Resources, Mining & Metals, Utilities or Power) – with hands on experience handling data driven...
-
Rueil-Malmaison, France BEICIP-FRANLAB Temps plein**Description offre**:**Objectives** Leading consulting firm in geoscience, guiding client decisions across all energy sectors, supported by research and development conducted by IFPEN Group, Beicip-Franlab is exploring the potential of AI throughout its entire value chain. Historically data-driven and utilizing machine learning techniques, Beicip-Franlab's...
-
Rueil-Malmaison, France IFP Energies nouvelles Temps plein**Detailed numerical simulations of porous media combustion using ammonia and hydrogen as fuels: multi-physics coupling and modelling**: - Réf **ABG-134952** - Sujet de Thèse - 09/01/2026 - Autre financement public - IFP Energies nouvelles - Lieu de travail- Rueil-Malmaison - Ile-de-France - France - Intitulé du sujet- Detailed numerical simulations of...
-
Stage Internship in Reservoir Geomechanics Modelling
il y a 2 semaines
Rueil-Malmaison, France BEICIP-FRANLAB Temps plein**Description offre**:**Objectives** As a leading consulting firm in geoscience and engineering challenges, supported by the research and innovation of the IFPEN Group, Beicip-Franlab continues to strengthen its expertise in advanced geomechanical modelling. In this context, Beicip-Franlab offers an internship dedicated to exploring innovative methodologies...
-
Finance Business Partner
il y a 1 semaine
Rueil-Malmaison, France Unilever Temps pleinWHY JOIN UNILEVER - To reach our ambition of doubling the size of our business and halving our environmental footprint, we need people with a winning mindset, a passion for consumers and an appetite to drive personal performance.A career at Unilever will see you join one of the world’s most attractive employers. We have gained our reputation as one of the...
-
Rueil-Malmaison, Île-de-France IFPEN Temps pleinData assimilation for predicting pollution maps under realistic environmental conditionsRéf ABG-134951Sujet de Thèse09/01/2026Contrat doctoralIFP Energies nouvellesLieu de travailRueil-Malmaison - Ile-de-France - FranceIntitulé du sujetData assimilation for predicting pollution maps under realistic environmental conditionsChamps scientifiquesPhysiqueMots...
