Maurice Kraus

Machine Learning Group, Computer Science Department, TU Darmstadt.
Hochschulstrasse 1, Building S1|03, Room 077, 64289 Darmstadt, Germany.
maurice (dot) kraus (at) cs (dot) tu-darmstadt (dot) de
mkraus.io

My research area focuses on explaining deep models in the context of multimodal time series data. Specifically, I work on techniques that enable humans to understand the reasoning behind the predictions made by machine learning models when they operate on complex and diverse time-varying data types, such as images, videos, audio, and text, that are collected over time.

Multimodal time series data is prevalent in various domains, such as healthcare, finance, and manufacturing, where it is often used for prediction tasks like disease diagnosis, stock price forecasting, and anomaly detection. My main goal is to develop techniques for visualizing and interpreting the contributions of each modality to the overall prediction, as well as methods that can explain the causal relationships among the different modalities and how they evolve over time. These techniques can help experts in various domains gain insight into the underlying processes that govern the data and the model's predictions, enabling them to make informed decisions based on the model's output.

Overall, my research has important implications for improving the trustworthiness and accountability of machine learning models so that they can be used reliably and effectively to solve real-world problems.

If you are interested in writing a thesis in the exciting field of deep learning and machine learning, or if you would like to work on a student project, please do not hesitate to contact me. I would be happy to discuss possible thesis or project topics with you, and I am also open to your own topic proposals that align with my research interests.