Menu

STUDENT SEMINAR 

 

Speaker:           Kushal Singh 

 

Topic:                Machine Learning Thermodynamics 

 

Date & Time:     Thursday, 18th March  2021 at 4:00 PM through MICROSOFT TEAMS 

 

https://teams.microsoft.com/l/meetup-join/19%3a95b3dfced9714083b3ea8ab65a1c6082%40thread.tacv2/1615876272807?context=%7b%22Tid%22%3a%226f15cd97-f6a7-41e3-b2c5-ad4193976476%22%2c%22Oid%22%3a%2230968044-0b08-4873-a9ee-481b8f662e29%22%7d

 

  Join conversation

teams.microsoft.com

 

Abstract:  

 

Machine learning is a paradigm whereby computer algorithms are designed to learn from and make predictions on data. Machine learning has been used to great success in many scientific applications such as predicting the structure of biomolecules1, locating specific binding regions2, identifying potential drug candidates3, searching for signatures of exotic particles from high energy collision data4, classifying phases of matter5 and finding the locations of phase transitions6.

 

In this seminar, I will be discussing how machine learning can be applied to thermodynamics. I will be focusing on two major applications – i) learning the underlying probability distribution of a Hamiltonian using a neural network architecture called Restricted Boltzmann Machine7, ii) using machine learning to predict the direction of the arrow of time for a given microscopic process8.

 

References: 

 

  1. Jumper J., Tunyasuvunakool, K., Kohli, P., Hassabis, D., and the AlphaFold Team, “Computational predictions of protein structures associated with COVID-19”, Version 3, DeepMind website (2020)
  2. Stepniewska-Dziubinska, M.M., Zielenkiewicz, P. & Siedlecki, P. “Improving detection of protein-ligand binding sites with 3D segmentation”, Sci Rep 10, 5035 (2020).
  3. Réda, C., Kaufmann, E. & Delahaye-Duriezade, A., “Machine learning applications in drug development”, Comput. Struct. Biotechnol. J. 18, 241-252(2020)
  4. Baldi, P., Sadowski, P. & Whiteson, D. “Searching for exotic particles in high-energy physics with deep learning”, Nat Commun 5, 4308 (2014)
  5. Carrasquilla, J. & Melko, R., “ Machine learning phases of matter”, Nature Phys 13, 431–434 (2017)
  6. Wetzel, S.J., “Unsupervised learning of phase transitions: From principal component analysis to variational autoencoders”, Phys. Rev. E, 96 022140 (2017)
  7. Torlai, G. & Melko, R.G., “Learning thermodynamics with Boltzmann machines”, Phys. Rev. B, 94, 165134 (2016)
  8. Seif, A., Hafezi, M. & Jarzynski, C., “Machine learning the thermodynamic arrow of time”, Nat. Phys. 17, 105–113 (2021)