Decentralized stochastic gradient langevin dynamics and hamiltonian monte carlo M Gürbüzbalaban, X Gao, Y Hu, L Zhu Journal of Machine Learning Research 22 (239), 1-69, 2021 | 18 | 2021 |
Fractional moment-preserving initialization schemes for training deep neural networks M Gurbuzbalaban, Y Hu International Conference on Artificial Intelligence and Statistics, 2233-2241, 2021 | 11 | 2021 |
Non-convex optimization via non-reversible stochastic gradient Langevin dynamics Y Hu, X Wang, X Gao, M Gurbuzbalaban, L Zhu arXiv preprint arXiv:2004.02823, 2020 | 11 | 2020 |
Fractional moment-preserving initialization schemes for training fullyconnected neural networks M Gürbüzbalaban, Y Hu arXiv preprint arXiv:2005.11878, 2020 | 5 | 2020 |
Heavy-tail phenomenon in decentralized sgd M Gurbuzbalaban, Y Hu, U Simsekli, K Yuan, L Zhu arXiv preprint arXiv:2205.06689, 2022 | 3 | 2022 |
Non-convex stochastic optimization via nonreversible stochastic gradient langevin dynamics Y Hu, X Wang, X Gao, M Gurbuzbalaban, L Zhu arXiv preprint arXiv:2004.02823, 2020 | 3 | 2020 |
Penalized Langevin and Hamiltonian Monte Carlo Algorithms for Constrained Sampling M Gürbüzbalaban, Y Hu, L Zhu arXiv preprint arXiv:2212.00570, 2022 | 1 | 2022 |
Cyclic and Randomized Stepsizes Invoke Heavier Tails in SGD than Constant Stepsize M Gürbüzbalaban, Y Hu, U Şimşekli, L Zhu arXiv preprint arXiv:2302.05516, 2023 | | 2023 |
Stochastic Gradient and Stochastic Gradient MCMC Methods for Bayesian Learning and Non-convex Optimization: Centralized and Decentralized Settings Y Hu Rutgers The State University of New Jersey, Graduate School-Newark, 2023 | | 2023 |
Cyclic and Randomized Stepsizes Invoke Heavier Tails in SGD. M Gürbüzbalaban, Y Hu, U Simsekli, L Zhu CoRR, 2023 | | 2023 |