Introduction to machine learning and its application to physics

From QCLab
Dr. Lee, Chang-Woo (KIAS)


Machine learning (ML) is now permeating into almost every field of science; physics seems to be no exception. Recently, many overseas groups are utilizing ML to attack hard-to-solve problems (mostly) in the context of many-body physics. Although concrete insight is still far from our grasp via this technique, ML often shows good performance beyond our expectation. Following the trend of using ML as a data-analyzing or feature-extracting tool for physics, this talk will briefly touch on the history, types, and background theories of ML; it will also browse through some recent studies in this line.