HEP Theory Journal Club

A High-Bias, Low-Variance introduction to Machine Learning

Omar Chammaa

McGill

In this talk, I will give an introduction to Machine Learning. I will state its goals, contrast it with classical statistics and discuss the infamous bias-variance tradeoff. Then, I will focus on the various optimization methods, the simplest of which is gradient descent and its various improvements like AdaGrad, ADAM and so on. Then I will show some examples of supervised machine learning algorithms. This talk is based on physics.comph-ph: 1803.08823v2.