| The convergence of sparsified gradient methods D Alistarh, T Hoefler, M Johansson, N Konstantinov, S Khirirat, C Renggli Advances in Neural Information Processing Systems 31, 2018 | 576 | 2018 |
| Mini-batch gradient descent: Faster convergence under data sparsity S Khirirat, HR Feyzmahdavian, M Johansson 2017 IEEE 56th Annual Conference on Decision and Control (CDC), 2880-2887, 2017 | 131 | 2017 |
| Distributed learning with compressed gradients S Khirirat, HR Feyzmahdavian, M Johansson arXiv preprint arXiv:1806.06573, 2018 | 94 | 2018 |
| Compressed gradient methods with hessian-aided error compensation S Khirirat, S Magnússon, M Johansson IEEE Transactions on Signal Processing 69, 998-1011, 2020 | 17 | 2020 |
| A flexible framework for communication-efficient machine learning S Khirirat, S Magnússon, A Aytekin, M Johansson Proceedings of the AAAI Conference on Artificial Intelligence 35 (9), 8101-8109, 2021 | 15 | 2021 |
| Gradient compression for communication-limited convex optimization S Khirirat, M Johansson, D Alistarh 2018 IEEE Conference on Decision and Control (CDC), 166-171, 2018 | 14 | 2018 |
| Communication efficient sparsification for large scale machine learning S Khirirat, S Magnússon, A Aytekin, M Johansson arXiv preprint arXiv:2003.06377, 2020 | 9 | 2020 |
| Improved step-size schedules for proximal noisy gradient methods S Khirirat, X Wang, S Magnússon, M Johansson IEEE Transactions on Signal Processing 71, 189-201, 2023 | 8 | 2023 |
| Clip21: Error feedback for gradient clipping S Khirirat, E Gorbunov, S Horváth, R Islamov, F Karray, P Richtárik arXiv preprint arXiv:2305.18929, 2023 | 4 | 2023 |
| On the Convergence of Federated Learning Algorithms without Data Similarity A Beikmohammadi, S Khirirat, S Magnússon arXiv preprint arXiv:2403.02347, 2024 | 3 | 2024 |
| Eco-Fedsplit: Federated learning with error-compensated compression S Khirirat, S Magnússon, M Johansson ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and …, 2022 | 3 | 2022 |
| Zeroth-order randomized subspace Newton methods E Berglund, S Khirirat, X Wang ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and …, 2022 | 3 | 2022 |
| Improved step-size schedules for noisy gradient methods S Khirirat, X Wang, S Magnússon, M Johansson ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and …, 2021 | 3 | 2021 |
| Compressed Federated Reinforcement Learning with a Generative Model A Beikmohammadi, S Khirirat, S Magnússon arXiv preprint arXiv:2404.10635, 2024 | 2 | 2024 |
| Distributed Momentum Methods Under Biased Gradient Estimations A Beikmohammadi, S Khirirat, S Magnússon arXiv preprint arXiv:2403.00853, 2024 | 2 | 2024 |
| Convergence bounds for compressed gradient methods with memory based error compensation S Khirirat, S Magnússon, M Johansson ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and …, 2019 | 2 | 2019 |
| Randomized first-order methods for convex optimization S Khirirat Master's thesis. KTH Royal Institute of Technoloay, Stockholm, 2016 | 2 | 2016 |
| Randomized first-order methods for convex optimization: Improved convergence rate bounds and experimental evaluations S Khirirat | 1 | 2016 |
| Balancing Privacy and Performance for Private Federated Learning Algorithms X Hou, S Khirirat, M Yaqub, S Horvath arXiv preprint arXiv:2304.05127, 2023 | | 2023 |
| Revisiting the Curvature-aided IAG: Improved Theory and Reduced Complexity E Berglund, S Khirirat, X Wu, S Magnússon, M Johansson IFAC-PapersOnLine 56 (2), 5221-5226, 2023 | | 2023 |
Mikael JohanssonProfessor of Electrical Engineering, KTH在 kth.se 的电子邮件经过验证
