Research • Reports | Announcements | Moodle
© Centre for Modeling & Simulation, Savitribai Phule Pune University
| Savitribai Phule Pune University |
| Title | Accelerating Materials Design Through Machine Learning |
| Speaker |
Sandip De,
Scientist EPFL, Lausanne, Switzerland
Dr. Sandip De is a Scientist at prestigious EPFL, Lausanne, Switzerland. He is expert in developing and using machine learning framework to unlock full potential of big databases in materials science for accelerating next generation materials design and discovery. He has extensive academic and industrial research experience in the field of quantum chemistry involving structure prediction and characterization of diverse class of materials using state of the art quantum mechanical, classical and multi-scale materials modelling. He has recently launched a machine learning based app to visualize molecular database. Apart from being scientist he is also a renowned photographer. Besides winning several photography awards from time to time, his works have been published in international media and journals like National Geographic. You can enjoy his work at Sandip De Photography. |
| Date & Time | Thursday, 08 March 2018 | 15:00-16:00 |
| Venue |
Kelkar Laboratory Centre for Modeling and Simulation, Savitribai Phule Pune University |
| Abstract | Advancement of technology requires discovery and design of new materials with targeted properties. General intuition based experimental efforts are not sufficient enough achieving the goal alone without the guidance of theory. Computational materials design promises to greatly accelerate the process. Several collaborative efforts across the world have been contributing to this goal by building databases of structures, containing between thousands and millions of distinct hypothetical compounds, along with their properties computed with highly accurate electronic-structure calculation methods eg. Density functional theory and Coupled-Cluster methods. These databases are providing interesting new ways to explore and learn structure-property relationship through machine learning framework. I will discuss how machine learning methods are being used to analyse these huge amounts of atomic structure data, as well as for exploring different ways of learning and predicting different key electronic structure properties at a fraction of a cost of the traditional quantum chemistry methods. |
| Organizer/Host | Bhalchandra Pujari, Centre for Modeling and Simulation, Savitribai Phule Pune University |
Want to know about the next programme? — Subscribe to our announcements list!