Student Seminar

Machine Learning: Recent Advances in Physics and Technology

Fri, 10 Nov 2017 1:30 PM
Print

Student Seminar
Machine Learning: Recent Advances in Physics and Technology

Jeonghun Lee
Dept of Physics, SFU

Nov 10, 2017 at 1:30pm in P8445.2

Synopsis

Artificial intelligence and machine learning have become widespread in press and scientific papers. Over the past decade, speech recognition, web search, and self-driving cars were made possible by machine learning which implements so-called ‘artificial neural networks’. The idea is to allow a computer system to classify and recognize patterns of enormous complexity as the human brain and make decisions and perform tasks without being explicitly programmed.

Recently, scientific communities have utilized machine learning in a vast number of applications. Hybridizing machine learning with Monte Carlo simulation, phase transitions and order parameter directly from raw state configurations can be identified in condensed matter systems. Astrophysicists and biologists have demonstrated using ‘convolutional neural networks’ that gravitational lens candidates used for dark matter research can be detected and the shape red blood cells on a patient blood can be classified which gives a new avenue for diagnosing sickle cell disease. In this seminar, I will go through these recent advances.

[1] J. Carrasquilla and R. G. Melko, 1 (2016).

[2] C. E. Petrillo, C. Tortora, S. Chatterjee, G. Vernardos, L. V. E. Koopmans, G. V. Kleijn, N. R. Napolitano, G. Covone, P. Schneider, A. Grado, and J. McFarland, Mon. Not. R. Astron. Soc. 472, 1129 (2017).

[3] M. Xu, D. P. Papageorgiou, S. Z. Abidi, M. Dao, H. Zhao, and G. E. Karniadakis, PLOS Comput. Biol. 13, e1005746 (2017).