Zalán Borsos

PhD Student
zalan.borsos@inf.ethz.ch
External Website
CAB F 65.2
+41 44 632 43 89
My research interests lie in developing data summarization methods that possess strong approximation guarantees. The main technique I employ is summarization via coresets. This approach relies on retaining a small, quintessential subset of the data by trading off slight degradations in performance for large speedup and memory gains.

2021
Semi-supervised Batch Active Learning via Bilevel Optimization Z. Borsos, M. Tagliasacchi, A. Krause In Proc. International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021 [bibtex] [abstract] [pdf] [code]
2020
Coresets via Bilevel Optimization for Continual Learning and Streaming Z. Borsos, M. Mutný, A. Krause In Proc. Neural Information Processing Systems (NeurIPS), 2020 [bibtex] [abstract] [pdf] [code]
2019
Online Variance Reduction with Mixtures Z. Borsos, S. Curi, K. Y. Levy, A. Krause In Proc. International Conference on Machine Learning (ICML), 2019 [bibtex] [abstract] [pdf]
2018
Online Variance Reduction for Stochastic Optimization Z. Borsos, A. Krause, K. Y. Levy In Proc. Conference on Learning Theory (COLT), 2018 [bibtex] [abstract] [pdf] [code]