ZITI’s Computer Systems Group (CSG) is part of a new center for “Model-based AI” funded by the Carl Zeiss Foundation. The main goal of this center is to use model-based artificial intelligence to improve the treatment of cancer. In this context, we will explore new methods for green deep learning, including model compression for the training phase as well as scheduling for heterogeneous environments.
Heidelberg University, Model-Based AI: Physical Models and Deep Learning for Imaging and Cancer Treatment
The establishment of a center on model-based artificial intelligence is intended to improve the treatment of cancer. It is investigated how pre-programmed knowledge on cancer research can support AI systems more efficiently.
How can knowledge from a specific topic area, such as cancer research, be implemented in Deep Learning methods to achieve more effective processing or problem solving? For these and similar questions, the project plans to establish a center on model-based artificial intelligence. The focus of the research project is on the concrete application in the treatment of cancer. An interdisciplinary team from the faculties of computer science, mathematics, physics and medicine is analyzing various issues in AI research, including the reliability of learning data, object recognition and the quality of data storage and evaluation.