Computational Laboratory for Energy And Nanoscience

University Homepage | Faculty of Science | Faculty of Graduate Studies | Registrar's Office |
Physics Homepage | Graduate Studies in Physics | Department contact information
Map it | City of Oshawa | Regional News | Local Weather | Government of Canada
subglobal4 link | subglobal4 link | subglobal4 link | subglobal4 link | subglobal4 link | subglobal4 link | subglobal4 link
subglobal5 link | subglobal5 link | subglobal5 link | subglobal5 link | subglobal5 link | subglobal5 link | subglobal5 link
subglobal6 link | subglobal6 link | subglobal6 link | subglobal6 link | subglobal6 link | subglobal6 link | subglobal6 link
subglobal7 link | subglobal7 link | subglobal7 link | subglobal7 link | subglobal7 link | subglobal7 link | subglobal7 link
subglobal8 link | subglobal8 link | subglobal8 link | subglobal8 link | subglobal8 link | subglobal8 link | subglobal8 link

Manuscript Summary - Deep neural networks for learning operators through observation: the case of the 2d Ising model

In this article, we show that deep neural networks (CNN to be exact) are able to learn how to compute properties of physical models through observation alone. You do not need to "teach" them equations or math. Here, we showed a CNN lots of examples of checkboard patterns, where each had a simple energy based on a "white likes white, black likes black" model. This is the same model we explored here, except now we used deep learning rather than conventational machine learning approaches. The deep learning model we trained is so good that it can be used to explore phase transitions. We also show that it is really efficient. For a more difficult energy (one based on charged particles which have long range interactions), we trained a DNN that was 3 orders of magnitude (i.e. 1000x) faster than a conventional calculation.

UOIT