I am a Research Fellow at Politecnico of Milano (Italy) working on Streaming Edge AI.

Selected publications

SMOTE-OB: Combining SMOTE and Online Bagging for Continuous Rebalancing of Evolving Data Streams

A Bernardo, E Della Valle. IEEE International Conference on Big Data (Big Data), 2021. DOI

A novel cost-sensitive ensemble strategy that uses Online Bagging and VFC-SMOTE to over/undersample the minority and majority classes. Access Paper

VFC-SMOTE: Very Fast Continuous Synthetic Minority Oversampling for Evolving Data Streams

A Bernardo, E Della Valle. Journal of Data Mining and Knowledge Discovery (DAMI) special issue at the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD), 2021. DOI

A new meta-strategy inspired by SMOTE, Borderline-SMOTE and Dynamic Multidimensional Histograms and, in addition to improve the classification performances w.r.t. the state-of-the-art methods, as C-SMOTE does, it focuses on improving the computational performances as time to execute or memory used, too. Access Paper

C-SMOTE: Continuous Synthetic Minority Oversampling for Evolving Data Streams

A Bernardo, H M Gomes, J Montiel, B Pfahringer, E Della Valle, A Bifet. IEEE International Conference on Big Data (Big Data), 2020. DOI

A new meta-strategy for rebalancing imbalanced data streams in case of concept drift. It uses one element at time to update the model and it continuously rebalances the stream using a modified version of the well known SMOTE technique. Access Paper