Deep Learning:

• Provable In-context Learning for Mixture of Linear Regressions using Transformers (with Yanhao Jin and Lifeng Lai), 2024.

• Transformers Handle Endogeneity in In-Context Linear Regression (with Haodong Liang and Lifeng Lai), 2024.

• From Stability to Chaos: Analyzing Gradient Descent Dynamics in Quadratic Regression (with Xuxing Chen, Promit Ghosal and Bhavya Agrawalla), Transactions on Machine Learning Research (TMLR), 2024.

• Gaussian random field approximation via Stein's method with applications to wide random neural networks (with Larry Goldstein, Nathan Ross and Adil Salim), Applied and Computational Harmonic Analysis, 2024. 

Sampling:

• Improved Finite-Particle Convergence Rates for Stein Variational Gradient Descent (Sayan Banerjee and Promit Ghosal), 2024.

• A Separation in Heavy-Tailed Sampling: Gaussian vs. Stable Oracles for Proximal Samplers (with Ye He, Alireza Mousavi-Hosseini and Murat A. Erdogdu), NeurIPS, 2024. 

• Regularized Stein Variational Gradient Flow  (with Ye He, Bharath K. Sriperumbudur and Jianfeng Lu), Foundations of Computational Mathematics, 2024.

• Towards Understanding the Dynamics of Gaussian-Stein Variational Gradient Descent (with Tianle Liu, Promit Ghosal and Natesh Pillai), NeurIPS, 2023.

• Mean-square Analysis of Discretized Itô Diffusions for Heavy-tailed Sampling (with Ye He, Tyler Farghly and Murat A. Erdogdu), Journal of Machine Learning Research, 2024.

• An analysis of Transformed Unadjusted Langevin Algorithm for Heavy-tailed Sampling (with Ye He and Murat A. Erdogdu), IEEE Information Theory, 2024.

• Forward-backward Gaussian variational inference via JKO in the Bures-Wasserstein Space (with Michael Diao, Sinho Chewi and Adil Salim), ICML, 2023.

• Towards a Complete Analysis of Langevin Monte Carlo: Beyond Poincaré Inequality (with Alireza Mousavi-Hosseini, Tyler Farghly, Ye He and Murat A. Erdogdu), COLT, 2023.

• Improved Discretization Analysis for Underdamped Langevin Monte Carlo (with Matthew Zhang, Sinho Chewi, Mufan Bill Li, and Murat A. Erdogdu),  COLT, 2023.

• Towards a Theory of Non-Log-Concave Sampling: First-Order Stationarity Guarantees for Langevin Monte Carlo (with Sinho Chewi, Murat A. Erdogdu, Adil Salim and Matthew Zhang), COLT, 2022.

• Stochastic Zeroth-order Discretizations of Langevin Diffusions for Bayesian Inference (with Abhishek Roy, Lingqing Shen and Saeed Ghadimi), Bernoulli [link], 2022.

• On the Ergodicity, Bias and Asymptotic Normality of Randomized Midpoint Sampling Method (with Ye He and Murat Erdogdu), NeurIPS, 2020.

Stochastic Optimization:

• High-dimensional scaling limits and fluctuations of online least-squares SGD with smooth covariance (with Promit Ghosal and Ye He), Annals of Applied Probability, 2024. 

• Stochastic Optimization Algorithms for Instrumental Variable Regression with Streaming Data (with Xuxing Chen, Abhishek Roy and Yifan Hu), NeurIPS, 2024.

• Zeroth-order Riemannian Averaging Stochastic Approximation Algorithms (with Jiaxiang Li and Shiqian Ma) SIAM Journal of Optimization, 2024.

• Online covariance estimation for stochastic gradient descent under Markovian sampling (with Abhishek Roy), Journal of the American Statistical Association (under major revision), 2024.

• Stochastic Nested Compositional Bi-level Optimization for Robust Feature Learning (with Xuxing Chen and Saeed Ghadimi), Mathematical Programming (under minor revision), 2024.

• Optimal algorithms for stochastic bilevel optimization under relaxed smoothness conditions (with Xuxing Chen and Tesi Xiao), Journal of Machine Learning Research, 2024.

• Statistical Inference for Linear Functionals of Online SGD in High-dimensional Linear Regression (with Bhavya Agrawalla and Promit Ghosal), 2023.

• Decentralized Stochastic Bilevel Optimization with Improved Per-Iteration Complexity (with Xuxing Chen, Minhui Huang and Shiqian Ma), ICML, 2023. 

• A One-Sample Decentralized Proximal Algorithm for Non-Convex Stochastic Composite Optimization (with Tesi Xiao, Xuxing Chen and Saeed Ghadimi), UAI, 2023.

• Stochastic Zeroth-order Riemannian Derivative Estimation and Optimization (with Jiaxiang Li and Shiqian Ma),  Mathematics of Operations Research [link], 2022. 

• Stochastic Zeroth-order Functional Constrained Optimization: Oracle Complexity and Applications (with Anthony Nguyen), INFORMS Journal on Optimization [link], 2022.

• Constrained Stochastic Nonconvex Optimization with State-dependent Markov Data (with Abhishek Roy and Saeed Ghadimi), NeurIPS, 2022. 

• A Projection-free Algorithm for Constrained Stochastic Multi-level Composition Optimization (with Tesi Xiao and Saeed Ghadimi), NeurIPS, 2022.

• Mirror Descent Strikes Again: Optimal Stochastic Convex Optimization under Infinite Noise Variance (with Nuri Mert Vural, Lu Yu, Stanislav Volgushev and Murat A. Erdogdu), COLT, 2022.

• Stochastic Multilevel Composition Optimization Algorithms with Level-Independent Convergence Rates (with Saeed Ghadimi and Anthony Nguyen), SIAM Journal on Optimization [link], 2022.

• Zeroth-order Nonconvex Stochastic Optimization: Handling Constraints, High-Dimensionality and Saddle-Points, (with Saeed Ghadimi), Foundations of Computational Mathematics [link], 2022. (Short version accepted in NeurIPS, 2018).

• Zeroth-Order Algorithms for Nonconvex Minimax Problems with Improved Complexities (with  Shiqian Ma, Meisam Razaviyayn and Zhongruo Wang), Journal of Global Optimization [link], 2022. 

• Stochastic Zeroth-Order Optimization under Nonstationarity and Nonconvexity (with Abhishek Roy, Saeed Ghadimi and Prasant Mohapatra), Journal of Machine Learning Research, 2022.

• Improved Complexities for Stochastic Conditional Gradient Methods under Interpolation-like Conditions (with Tesi Xiao and Saeed Ghadimi), Operations Research Letters [link], 2022.

• High-Probability Bounds for Robust Stochastic Frank-Wolfe Algorithm (with Tongyi Tang and Thomas C.M. Lee) UAI, 2022. 

• Statistical Inference for Polyak-Ruppert Averaged Zeroth-order Stochastic Gradient Algorithm (with Yanhao Jin and Tesi Xiao), 2021. 

• An Analysis of Constant Step Size SGD in the Non-convex Regime: Asymptotic Normality and Bias (with Lu Yu, Stanislav Volgushev and Murat A. Erdogdu), NeurIPS, 2021.

• Escaping Saddle-Points Faster under Interpolation-like Conditions (with Abhishek Roy,  Saeed Ghadimi and Prasant Mohapatra), NeurIPS, 2020.

• Normal Approximation for Stochastic Gradient Descent via Non-Asymptotic Rates of Martingale CLT, (with Andreas Anastasiou and Murat A. Erdogdu), COLT, 2019. 

• Online and Bandit Algorithms for Nonstationary Stochastic Saddle-Point Optimization (with Yifang Chen and Abhishek Roy), 2019.

Nonparametric, Geometric and Topological Statistics:

• Statistical-Computational Trade-offs for Greedy Recursive Partitioning Estimators (with Yan Shuo Tan and Jason M. Klusowski), 2024.

• Multivariate Gaussian Approximation for Random Forest via Region-based Stabilization (with Zhaoyang Shi, Chinmoy Bhattacharjee and Wolfgang Polonik), 2024

• Minimax Optimal Goodness-of-Fit Testing with Kernel Stein Discrepancy (with Omar Hagrass and Bharath Sriperumbudur), 2024. 

• Nonsmooth Nonparametric Regression via Fractional Laplacian Eigenmaps (with Zhaoyang Shi and Wolfgang Polonik), 2024. 

• Adaptive and non-adaptive minimax rates for weighted Laplacian-eigenmap based nonparametric regression (with Zhaoyang Shi and Wolfgang Polonik), International Conference on Artificial Intelligence and Statistics (AISTATS), 2024. 

• A Flexible Approach for Normal Approximation of Geometric and Topological Statistics (with Zhaoyang Shi and Wolfgang Polonik), Bernoulli, 2024.

• Functional linear and single-index models: A unified approach via Gaussian Stein identity (with Hans-Georg Müller and Bharath Sriperumbudur), Bernoulli, 2024. 

• Topologically penalized regression on manifolds (with Olympio Hacquard, Gilles Blanchard, Clément Levrard and Wolfgang Polonik),  Journal of Machine Learning Research, 2022. 

• Fractal Gaussian Networks: A sparse random graph model based on Gaussian Multiplicative Chaos (with Subhroshekar Ghosh and Xiaochuan Yang),  IEEE Transactions on Information Theory [link], 2022. (short version appeared in ICML 2020).

• Nonparametric Modeling of Higher-Order Interactions via Hypergraphons,  Journal of Machine Learning Research, 2021.  

• On the Optimality of Kernel-Embedding Based Goodness-of-Fit Test, (with  Tong Li and Ming Yuan), Journal of Machine Learning Research, 2021. 

• Ultrahigh Dimensional Feature Screening via RKHS Embeddings, (with Bharath  Sriperumbudur and Guy Lebanon), International Conference on Artificial Intelligence and Statistics (AISTATS), 2013.   

High-dimensional Statistics and Machine Learning:

• Meta-Learning with Generalized Ridge Regression: High-dimensional Asymptotics, Optimality and Hyper-covariance Estimation (with Yanhao Jin and Debashis Paul), 2024. 

• On Empirical Risk Minimization with Dependent and Heavy-tailed Data (with Abhishek Roy and Murat A. Erdogdu), NeurIPS, 2021. 

• Tensor Methods for Additive Index Models under Discordance and Heterogeneity, (with Jianqing Fan and Zhuoran Yang), Under Revision, 2018.

• Inhomogeneous Random Tensors and its Applications to Tensor PCA and Tensor Sparsification (with Subhroshekar Ghosh), 2020.

• Estimating High-dimensional Non-Gaussian Multiple Index Models via Stein’s Lemma,  (with Zhuoran Yang, Zhaoran Wang and  Han Liu), NeurIPS 2017.

• High-dimensional Non-Gaussian Single Index Models via Thresholded Score Function Estimation, (with Zhuoran Yang and  Han Liu), ICML 2017.

• Discussion: Estimating Structured High-Dimensional Covariance and Precision Matrices: Optimal Rates and Adaptive Estimation, (with  Ming Yuan), Electronic Journal of Statistics, 2016. 

• Smooth Sparse Coding via Marginal Regression for Learning Sparse Representations, (with Kai Yu  and Guy Lebanon),  Artificial Intelligence, 2016. Short version: International Conference on Machine Learning (ICML), 2013. (Best Paper Runner-up Award).

• High-dimensional Joint Sparsity Random Effects Model for Multi-task Learning, (with Kai Yu and Tong Zhang), Conference on Uncertainty in Artificial Intelligence (UAI), 2013.   

• A Landmark Selection Approach for Multi-output Prediction Problems, (with Guy Lebanon),  International Conference on Machine Learning  (ICML), 2012.   

• Unsupervised Supervised Learning II: Margin-Based Classification without Labels, (with Pinar  Donmez  and Guy Lebanon), Journal of Machine Learning Research, 12, pp. 3119-3145, 2011. Short version:  International Conference on Artificial Intelligence and Statistics  (AISTATS), 2011.   

• Unsupervised Supervised Learning I: Estimating Classification and Regression Errors without Labels, (with Pinar  Donmez and Guy Lebanon), Journal of Machine Learning Research,  2010.  

• Asymptotic Analysis of Generative Semi-Supervised Learning, (with Joshua  Dillon and Guy Lebanon), International Conference on Machine Learning  (ICML), 2010.  

• Dimensionality Reduction for Text using Domain Knowledge, (with Yi Mao and Guy Lebanon), International Conference on Computational Linguistics  (COLING), 2010.  

• A Fast Algorithm for Nonnegative Tensor Factorization using Block Coordinate Descent  and an Active-set-type Method,  (with Jingu Kim,  Andrey Puretskiy,  Micheal Berry  and Haesun Park),  Text Mining Workshop: SIAM International Conference on Data Mining , 2010.