| Newton-type methods for non-convex optimization under inexact Hessian information P Xu, F Roosta, MW Mahoney Mathematical Programming 184 (1-2), 35-70, 2020 | 191 | 2020 |
| Second-order optimization for non-convex machine learning: An empirical study P Xu, F Roosta, MW Mahoney Proceedings of the 2020 SIAM International Conference on Data Mining, 199-207, 2020 | 147 | 2020 |
| Sub-sampled newton methods ii: Local convergence rates F Roosta-Khorasani, MW Mahoney arXiv preprint arXiv:1601.04738, 2016 | 123 | 2016 |
| Sub-sampled Newton methods with non-uniform sampling P Xu, J Yang, F Roosta, C Ré, MW Mahoney Advances in Neural Information Processing Systems 29, 2016 | 121 | 2016 |
| Improved bounds on sample size for implicit matrix trace estimators F Roosta-Khorasani, U Ascher Foundations of Computational Mathematics 15 (5), 1187-1212, 2015 | 116 | 2015 |
| Sub-sampled Newton methods F Roosta-Khorasani, MW Mahoney Mathematical Programming 174, 293-326, 2019 | 114 | 2019 |
| Giant: Globally improved approximate newton method for distributed optimization S Wang, F Roosta, P Xu, MW Mahoney Advances in Neural Information Processing Systems 31, 2018 | 113 | 2018 |
| Parallel local graph clustering J Shun, F Roosta-Khorasani, K Fountoulakis, MW Mahoney arXiv preprint arXiv:1604.07515, 2016 | 81 | 2016 |
| Stochastic algorithms for inverse problems involving PDEs and many measurements F Roosta-Khorasani, K Van Den Doel, U Ascher SIAM Journal on Scientific Computing 36 (5), S3-S22, 2014 | 51 | 2014 |
| DINGO: Distributed Newton-type method for gradient-norm optimization R Crane, F Roosta Advances in Neural Information Processing Systems 32, 2019 | 41 | 2019 |
| Inexact non-convex Newton-type methods Z Yao, P Xu, F Roosta-Khorasani, MW Mahoney arXiv preprint arXiv:1802.06925, 2018 | 37 | 2018 |
| Newton-MR: Inexact Newton Method with minimum residual sub-problem solver F Roosta, Y Liu, P Xu, MW Mahoney EURO Journal on Computational Optimization 10, 100035, 2022 | 33* | 2022 |
| Invariance of weight distributions in rectified MLPs R Tsuchida, F Roosta, M Gallagher International Conference on Machine Learning, 4995-5004, 2018 | 29 | 2018 |
| GPU accelerated sub-sampled Newton's method for convex classification problems S Kylasa, F Roosta, MW Mahoney, A Grama Proceedings of the 2019 SIAM international conference on data mining, 702-710, 2019 | 28 | 2019 |
| The reproducing Stein kernel approach for post-hoc corrected sampling L Hodgkinson, R Salomone, F Roosta arXiv preprint arXiv:2001.09266, 2020 | 24 | 2020 |
| Assessing stochastic algorithms for large scale nonlinear least squares problems using extremal probabilities of linear combinations of gamma random variables F Roosta-Khorasani, GJ Székely, UM Ascher SIAM/ASA Journal on Uncertainty Quantification 3 (1), 61-90, 2015 | 24 | 2015 |
| Data completion and stochastic algorithms for PDE inversion problems with many measurements F Roosta-Khorasani, K Doel, U Ascher arXiv preprint arXiv:1312.0707, 2013 | 24 | 2013 |
| Optimization methods for inverse problems N Ye, F Roosta-Khorasani, T Cui 2017 MATRIX Annals, 121-140, 2019 | 21 | 2019 |
| Stochastic normalizing flows L Hodgkinson, C van der Heide, F Roosta, MW Mahoney arXiv preprint arXiv:2002.09547, 2020 | 20 | 2020 |
| Variational perspective on local graph clustering K Fountoulakis, F Roosta-Khorasani, J Shun, X Cheng, MW Mahoney Mathematical Programming 174, 553-573, 2019 | 20 | 2019 |
Michael MahoneyProfessor of Statistics, UC BerkeleyVerified email at stat.berkeley.edu
