Geoffrey Decrouez

These lecture notes (.pdf ) are the result of four years of teaching at the Higher School of Economics (2014-2018), and in 2018, at the Yandex School of Data Analysis. They are based on various textbooks, notes from colleagues, mini-courses, and personal notes, whose references can be found at the end of each section (exception for the Causal Inference section, which I started writing after 2018). The aim is to give the reader an overview of the theoretical fundamentals underpinning modern techniques in machine learning. A brief but formal account of probability theory and mathematical statistics is first given, in order to appreciate in greater details the mathematics involved in modern algorithmic techniques. When relevant,  we provide some additional exercises (pbs).