We tackle unique challenges ranging from exploring image patterns to the intricate analysis of complex biological structures such as proteins. We aim to redefine the boundaries of effective and efficient processing of large datasets by exploiting machine learning techniques in data organizations. As part of the DISA laboratory and in close collaboration with the CERIT-SC center, we aim to discover patterns within vast amounts of complex data.
The basic search project Learned Indexing for Similarity Searching investigates an alternative to the prevalent paradigm of creating a hierarchical structure according to mutual distances between data items. We focus on utilizing machine learning models to replace decisions based on pivots, thus posing similarity search as a classification problem.
This research is led by Vlastislav Dohnal and Matej Antol.
More information is here.
