Heidelberg University

Machine learning in science and industry

Alex Rogozhnikov and Tatiana Likhomanenko, Yandex, Moscow

Abstract:

The success of machine learning methods became so obvious in the past two years, that nowadays media refers to them as 'artificial intelligence'. Today this 'AI' demonstrates impressive results in voice and facial recognition available in our smartphones, recommends music, books, and videos, predicts traffic jams and weather. At the same time, machine learning has numerous lesser known applications in fundamental science, for instance, in High Energy Physics and Astrophysics.

In this course we discuss different real-world problems and the ways to solve them efficiently with various machine learning techniques: we will start from the algorithms developed in the 20th century and proceed to the modern ones, including gradient boosting and deep learning.

We offer to the participants the possibility to apply and test some of the discussed methods in optional practical hands-on exercises.