Summary
Marco Frasca is a computer science researcher and associate professor at Università degli Studi di Milano with 11 years of experience building machine learning methods for bioinformatics and beyond. His work spans cost-sensitive and multitask classification for highly unbalanced problems, novel parametric Hopfield network models, and applications in gene function prediction, drug repositioning and phenotype prioritization. Recently he has focused on practical deep learning model compression—pruning, quantization and low-rank factorization—to design multi-criteria data structures and learned indexes. He blends theoretical modeling with applied biology problems, having moved from chromatin-based gene expression prediction to scalable learned systems. Based in Milan, he serves on the university’s PhD board and is known for tackling class-imbalance and task-dissimilarity challenges that often go overlooked in standard ML pipelines.
10 years of coding experience
6 years of employment as a software developer
University of Milan