High-dimensional and Functional Data Analysis
In the realm of Artificial Intelligence (AI), the group is exploring ways to address a critical question: how can we manage multiple tests effectively using less complex methods like lasso, slope, and neural networks (NN)?
Specifically, they are investigating techniques to control the False Discovery Rate (FDR) while utilizing NN. This involves the practical development of algorithms for regularization methods. Often, these testing procedures are integrated with AI algorithms, which adds an extra layer of computational complexity. Gaussian processes, which are infinite-dimensional, represent a strong tradition within the Department. Currently, these processes are being harnessed to comprehend NN and are also finding broader applications in the field of AI. Despite their general nature, these methods have practical applications in diverse fields such as epidemiology and genomics.
Researchers
- Krzysztof Podgorski
- Jonas Wallin
- Malgorzata Bogdan
- Iuliana Ionita-Laza
- a newly hired guest professor.