401-3620-19L  Student Seminar in Statistics: Adversarial and Robust Machine Learning

SemesterSpring Semester 2019
LecturersP. L. Bühlmann, M. H. Maathuis, N. Meinshausen, S. van de Geer
Periodicityevery semester recurring course
Language of instructionEnglish
CommentNumber of participants limited to 22.

Mainly for students from the Mathematics Bachelor and Master Programmes who, in addition to the introductory course unit 401-2604-00L Probability and Statistics, have heard at least one core or elective course in statistics. Also offered in the Master Programmes Statistics resp. Data Science.


AbstractAs statistical and machine learning models are increasingly employed in many real-world applications it becomes more important to understand the vulnerabilities and robustness properties of these models.
In the first part of this seminar, we will study papers relating to adversarial examples. In the second part of the course, we will review other types of distribution shifts.
Learning objectiveAfter this seminar, you should know
- properties of adversarial examples
- some attacks and defenses
- some concepts from robust optimization and distributional robustness
- other distribution shifts that can fool machine learning models in general and neural networks in particular
ContentAs statistical and machine learning models are increasingly employed in many real-world applications it becomes more important to understand the vulnerabilities and robustness properties of these models. In the first part of this seminar, we will study papers relating to adversarial examples, covering their properties, various attacks and defenses. In the second part of the course, we will review other types of distribution shifts, posing significant challenges for state-of-the-art machine learning models. Some parts of the seminar will be devoted to implementing these methods in python.
Prerequisites / NoticeWe require at least one course in statistics or machine learning and basic knowledge in computer programming. Some background knowledge in deep learning is helpful but not strictly required.
Topics will be assigned during the first meeting.