© Fotolia / maciek905

Introduction to machine learning

Machine learning and artificial intelligence now play a role in many areas of science, from astronomy to psychiatry. The interdisciplinary lecture is intended to serve as a first introduction to the methods of machine learning and their application in science.

Learnwebsite of the course

  • Dates and Room

    Block Course

    16.09.2024 - 20.09.2024

    in room  KP/TP 304
     

    Schedule Mo- Fr
    9.15 - 10:45 Lecture
    10:45 - 11:15 Break
    11:15 - 12:00 Lecture
    12:00 - 12:15 Pause
    12:15-13:00 Lecture

     

     

  • Content, literature and requirements

    Content of the lecture

    • Introduction to machine learning - general concepts, terms, challenges and problems
    • Regression, Naive-Bayes, Support Vector Machines, Principal Component analysis, Clustering, ...
    • Neural networks and deep learning
    • Reinforcement learning
    • Applying machine learning with Python to examples from different areas of science

     

    Literature

    The lecture is mainly based on the following sources (don't panic, even if the word physics appears in the first reference)

    • A. Géron, Hands-On Machine Learning with Scikit-Learn and TensorFlow: Concepts, Tools, and Techniques for Building Intelligent Systems
    • M. Nielsen, Neural Networks and Deep Learning, http: //neuralnetworksanddeeplearning.com
    • Christopher M. Bishop, Pattern Recognition and Machine Learning
    • R. S. Sutton and A. G. Barto, Reinforcement Learning: An Introduction
    • A. Burkov, The Hundred-Page Machine Learning Book
    • P. Mehta et. al., A high-bias, low-variance introduction to Machine Learning for physicists, arXiv: 1803.08823 (2019)

    The lecture gives a more comprehensive overview of the literaur.

     

    Requirements

    To understand the algorithms

    • A little bit of linear algebra
    • Derivatives in one and more dimensions

    To implement the methods

    • Elementary programming skills in Python

    The following libraries are used in the implementation (available for all operating systems - will be updated)

    Alternatively (everything online, i.e. nothing needs to be installed + large computer resources)

    Kaggle