We are looking for a motivated student to join a Project on the intersection of Physics and Machine Learning. The position is in the Materials Simulation & Modelling (MSM) Group in the Applied Physics Department (TU/e). MSM is a group dedicated to the simulation and design of novel materials and complex molecules for innovative, exciting and groundbreaking energy applications. This PhD position will be to work in the project entitled: 'Towards in silico DNA for materials', supported by three research institutes at TU/e, ICMS, EIRES, and EAISI. Exploring this field of research with the three institutes will contribute to a strong TU/e expert network in this area. Jointly we strive to push the boundaries of science. The project will be carried out under the supervision of prof. Sofia Calero and co-supervision of Dr. Vlado Menkovski. Prof. Calero is an expert in simulation and modelling from the Applied Physics department at TU/e . Dr. Vlado Menkovsk is an expert in Machine Learning from the Mathematics and Computer Science department at TU/e.
PhD programs at TU/e are four-year research positions, having as a primary goal to educate excellent, independent researchers. The program is in English and entails post-master level education in the form of courses and projects aiming at cutting-edge research that results in scientific publications and specific practical applications. For more information about the TU/e PhD program https://www.tue.nl/en/education/graduate-school/phd-at-tue/
We are developing a novel framework for Materials Discovery based on a combination of first principle Physics simulation and Machine Learning based simulation of materials. In the scope of this project we are also specifically looking to identify the most efficient zeolite for carbon capture. We aim to use Machine Learning based simulation, relying on Deep Generartive Models to significantly scale-up the simulation of materials that would allow us to efficiently explore the parameter space of the materials to uncover the materials with desired properties.
In the context of Material Science we aim to advance the state of the art by developing novel methods for discovery of new materials. As a multidisciplinary project we also aim to advance the state of the art of Machine Learning. Driven by the application of Deep Generatrive models for simulation of materials, we aim to develop new methodologies that are well suited for this task.
We will publish in top venues of Material and Computer Science (see details about publicatons and journals at www.tue.nl/msm
as well as Machine Learning venues as ICML, NeurIPS, ICLR and AAAI.