The project is developing novel quantum reservoir computing methods for the analysis of complex biomedical data. In collaboration with Merck, LMU is investigating the conditions under which QRC can offer advantages over classical methods and is developing new approaches for interpretable AI in personalized medicine and drug discovery.
The project aims to further develop quantum reservoir computing (QRC) methods and make them ready for use in the analysis of complex biomedical data. The focus is on the systematic investigation and quantification of the quantum advantage of QRC methods compared to classical methods – especially for small, difficult-to-interpret data sets, such as those typically found in drug discovery and personalized medicine.
LMU and Merck are developing new QRC architectures that are specifically optimized for small, high-dimensional data sets, integrating biomedical domain knowledge directly into the model structure. A key innovation goal is the introduction of a novel metric for quantifying reservoir utilization, which allows for a robust assessment of model complexity. In addition, methods for improving the interpretability of QRC results through visualization techniques such as UMAP are being co-developed, and new generative QRC approaches for molecular structure prediction are being co-researched.