Keywords
protein-protein interaction prediction; deep learning; topological data analysis; anomaly detection; bioinformatics
Health related-scientific domains
Life Sciences related to Health
Applied Medical Technologies, Diagnostics, Therapies
Research expertise/competencies
My PhD thesis focused on studying representation learning methods in the context of classification tasks with noisy labels. In particular, I investigated applications of autoencoders and contrastive learning in protein-protein interaction prediction. Previously, I worked on anomaly detection in medical images. I am currently interested in applying methods based on topological data analysis in biomedical tasks.