Data Science

Selected Publications

• Thanh Dang, Melih Barsbey, AKM Rokonuzzaman Sonet, Mert Gurbuzbalaban, Umut Simsekli and Lingjiong Zhu "Algorithmic stability of stochastic gradient descent with momentum under heavy-tailed noise" (2025). arXiv: 2502.00885.

• Hengrong Du, Qi Feng, Changwei Tu, Xiaoyu Wang and Lingjiong Zhu "Non-reversible Langevin algorithms for constrained sampling" (2025). arXiv:2501.11743.

• Mert Gurbuzbalaban, Mohammad Rafiqul Islam, Xiaoyu Wang and Lingjiong Zhu "Generalized EXTRA stochastic gradient Langevin dynamics" (2024). arXiv:2412.01993.

• George Rapakoulias, Ali Reza Pedram, Fengjiao Liu, Lingjiong Zhu and Panagiotis Tsiotras "Go with the flow: Fast diffusion for Gaussian mixture models" (2024). arXiv:2412.09059.

• Xuefeng Gao and Lingjiong Zhu "Convergence analysis for general probability flow ODEs of diffusion models in Wasserstein distances" (2025). International Conference on Artificial Intelligence and Statistics (AISTATS 2025).

• Umut Simsekli, Mert Gurbuzbalaban, Sinan Yildirim and Lingjiong Zhu "Privacy of SGD under Gaussian or heavy-tailed noise: Guarantees without gradient clipping" (2024). arXiv:2403.02051.

• Xuefeng Gao, Hoang M Nguyen and Lingjiong Zhu "Wasserstein convergence guarantees for a general class of score-based generative models" (2025). Journal of Machine Learning Research. 26, 1-54.

• Mert Gurbuzbalaban, Yuanhan Hu, Umut Simsekli, Kun Yuan and Lingjiong Zhu "Heavy-tail phenoemon in decentralized SGD" (2025). IISE Transactions. 57, 788-802.

• Mert Gurbuzbalaban, Yuanhan Hu and Lingjiong Zhu "Penalized overdamped and underdamped Langevin Monte Carlo algorithms for constrained sampling" (2024). Journal of Machine Learning Research. 25, 1-67.

• Lingjiong Zhu, Mert Gurbuzbalaban, Anant Raj and Umut Simsekli "Uniform-in-time Wasserstein stability bounds for (noisy) stochastic gradient descent" (2023). Advances in Neural Information Processing Systems. (NeurIPS 2023).

• Anant Raj, Lingjiong Zhu, Mert Gurbuzbalaban and Umut Simsekli "Algorithmic stability of heavy-tailed SGD with general loss functions" (2023). International Conference on Machine Learning.

• Anant Raj, Melih Barsbey, Mert Gurbuzbalaban, Lingjiong Zhu and Umut Simsekli "Algorithmic stability of heavy-tailed stochastic gradient descent on least squares" (2023). International Conference on Algorithmic Learning Theory.

• Mert Gurbuzbalaban, Yuanhan Hu, Umut Simsekli and Lingjiong Zhu "Cyclic and randomized stepsizes invoke heavier tails in SGD than constant stepsizes" (2023). Transactions on Machine Learning Research.

• Xuefeng Gao, Mert Gurbuzbalaban and Lingjiong Zhu "Global convergence of stochastic gradient Hamiltonian Monte Carlo for non-convex stochastic optimization: Non-asymptotic performance bounds and momentum-based acceleration" (2022). Operations Research. 70, 2931-2947.

• Alireza Fallah, Mert Gurbuzbalaban, Asuman Ozdaglar, Umut Simsekli and Lingjiong Zhu "Robust distributed accelerated stochastic gradient methods for multi-agent networks" (2022). Journal of Machine Learning Research. 23, 1-96.

• Mert Gurbuzbalaban, Umut Simsekli and Lingjiong Zhu "The heavy-tail phenomenon in SGD" (2021). International Conference on Machine Learning.

• Alexander Camuto, Xiaoyu Wang, Lingjiong Zhu, Mert Gurbuzbalaban, Chris Holmes and Umut Simsekli "Asymmetric heavy tails and implicit bias in Gaussian noise injections" (2021). International Conference on Machine Learning. 

• Alexander Camuto, George Deligiannidis, Murat Erdogdu, Mert Gurbuzbalaban, Umut Simsekli and Lingjiong Zhu "Fractal structure and generalization properties of stochastic optimization algorithms" Advances in Neural Information Processing Systems. (NeurIPS 2021).

• Hongjian Wang, Mert Gurbuzbalaban, Lingjiong Zhu, Umut Simsekli and Murat Erdogdu "Convergence rates of stochastic gradient descent under infinite noise variance" Advances in Neural Information Processing Systems. (NeurIPS 2021).

• Mert Gurbuzbalaban, Xuefeng Gao, Yuanhan Hu and Lingjiong Zhu "Decentralized stochastic gradient Langevin dynamics and Hamiltonian Monte Carlo" (2021). Journal of Machine Learning Research. 22, 1-69.

• Umut Simsekli, Lingjiong Zhu, Yee Whye Teh and Mert Gurbuzbalaban "Fractional underdamped Langevin dynamics: Retargeting SGD with momentum under heavy-tailed gradient noise" (2020). International Conference on Machine Learning.

• Xuefeng Gao, Mert Gurbuzbalaban and Lingjiong Zhu "Breaking reversibility accelerates Langevin dynamics for global non-convex optimization" (2020). Advances in Neural Information Processing Systems. (NeurIPS 2020).

• Bugra Can, Mert Gurbuzbalaban and Lingjiong Zhu "Accelerated linear convergence of stochastic momentum methods in Wasserstein distances" (2019). International Conference on Machine Learning, 891-901.