Welcome to the website of the Chair of Reliable Machine Learning

Machine Learning Laptop — © colourbox.de

The research group "Reliable Machine Learning" studies the properties of machine learning algorithms.
In view of the recent success of deep learning methods in applications like image recognition, speech recognition, and automatic translation, the group especially focuses on properties of deep neural networks.

Although a neural network trained e.g. for an image classification task might work well on "real inputs", it has been repeatedly shown empirically that such networks are vulnerable to adversarial examples:
a minimal perturbation (impercetible to a human) of the input data can cause the network to misclassify the input.
Thus, an important research area of the group is to mathematically understand the reasons for the existence of such adversarial examples (i.e., the instability of trained neural networks),
and - building on that understanding - to develop improved methods that yield provably robust neural networks.

The research group is supported by the Emmy Noether project "Stability and Solvability in Deep Learning".

Content of Chair of reliable machine learning

About us

The Team

Publications

Contact

Math News

05-03-2023

TRR workshop on scale interactions, data-driven modeling, and uncertainty in weather and climate

The workshop on scale interactions, data-driven modeling, and uncertainty in weather and climate was jointly organized by the CRC 181 “Energy transfers in Atmosphere and Ocean” and W2W. It took place at the Mathematical Institute for Machine Learning and Data Science in Ingolstadt from 27-30 March 2023.

The workshop was attended by about 80 participants from both CRCs, as well as by international scientists from the US, Italy and the UK (ca. 50 in person and ca. 30-50 online from all over the world). The topics covered were: uncertainty quantification and predictability, parametrizations and structure-preserving and invariant-conserving schemes, data-driven modeling and machine learning, data assimilation, waves in atmosphere and ocean, as well as wave-vortex interactions. Guest speakers included Rosimar Rios-Berrios (NCAR), Ted Shepherd (Univ. Reading) and Michael Gil (ENS, UCLA). The Early Career Scientists were active participants, e.g., they chaired the sessions.

The poster session on the first evening was very lively and the different communities (weather, climate, mathematics, ocean science, atmospheric science) met and discussed common methods and challenges, in an attempt to link recent advances in these areas and present new developments in the underlying theories, methods, and parameterizations. The poster session took place in a historical building of Hohe Schule Ingolstadt which served as a main building from 1503 till 1800 of first University in Bavaria (later LMU).

In addition to the poster session, the participants had plenty of occasions to exchange about their results during coffee breaks, as well as during social events that included the conference dinner and a visit of the medical museum. To those who are familier with the book from Mary Shelley will remember that Victor Frankenstein’s studied in Ingolstadt at University of Natural Sciences.

The slides and videos of the presentations are available here.

Childcare was organized during the meeting for two children of participants.

Text by Dr. Audine Laurian (LMU)

MIDS Seminar

Mathematical Institute for Machine Learning and Data Science

The Chair of Reliable Maschine Learning is part of the Mathematical Institute for Machine Learning and Data Science, MIDS.
Learn more about MIDS here.