Sample Publications (Show All)

• Online Bilevel Optimization: Regret Analysis of Online Alternating Gradient Methods
  (with Parvin Nazari, Bojian Hou, Li Shen, and Laura Balzano)
  International Conference on Artificial Intelligence and Statistics (AISTATS), 2024.
  Code Poster
• Fairness-Aware Estimation of Graphical Models
  (with Zhuoping Zhou, Bojian Hou, Qi Long, and Li Shen)
  Neural Information Processing Systems (NeurIPS), 2024.
  Code
• Transformers as Support Vector Machines
  (with Yingcong Li, Christos Thrampoulidis, and Samet Oymak)
  Preliminary Version at NeurIPS Workshop on Mathematics of Modern Machine Learning, 2023.
  Code Slide Poster
• Max-Margin Token Selection in Attention Mechanism
  (with Yingcong Li, Xuechen Zhang, and Samet Oymak)
  Neural Information Processing Systems (NeurIPS), 2023.
  Code Slide
  Spotlight (≈3% Acceptance)
• Fair Canonical Correlation Analysis
  (with Zhuoping Zhou, Bojian Hou, Boning Tong, Jia Xu, Yanbo Feng, Qi Long, and Li Shen)
  Neural Information Processing Systems (NeurIPS), 2023.
  Code Slide Poster
• FedNest: Federated Bilevel, Minimax, and Compositional Optimization
  (with Mingchen Li, Christos Thrampoulidis, and Samet Oymak)
  International Conference on Machine Learning (ICML), 2022.
  Code Slide
  Oral (≈2% Acceptance)
• Regularized and Smooth Double Core Tensor Factorization for Heterogeneous Data
  (with George Michailidis)
  Journal of Machine Learning Research 21, no. 23, 2022.
  Code
• Estimation of Graphical Models through Structured Norm Minimization
  (with George Michailidis)
  Journal of Machine Learning Research 18, no. 1, 2018.

All Papers (Show Sample)

2024

• Optimizing Attention with Mirror Descent: Generalized Max-Margin Token Selection
  (with Aaron Alvarado Kristanto Julistiono and Navid Azizan)
  Preliminary Version at NeurIPS Workshop on Mathematics of Modern Machine Learning, 2024.
  Code
• Fairness-Aware Estimation of Graphical Models
  (with Zhuoping Zhou, Bojian Hou, Qi Long, and Li Shen)
  Neural Information Processing Systems (NeurIPS), 2024.
  Code
• Online Bilevel Optimization: Regret Analysis of Online Alternating Gradient Methods
  (with Parvin Nazari, Bojian Hou, Li Shen, and Laura Balzano)
  International Conference on Artificial Intelligence and Statistics (AISTATS), 2024.
  Code Poster
• A Penalty-Based Method for Communication-Efficient Decentralized Bilevel Programming
  (with Parvin Nazari, Ahmad Mousavi, and George Michailidis)
  Automatica, 2024.
  

2023

• Transformers as Support Vector Machines
  (with Yingcong Li, Christos Thrampoulidis, and Samet Oymak)
  Preliminary Version at NeurIPS Workshop on Mathematics of Modern Machine Learning, 2023.
  Code Slide Poster
• Max-Margin Token Selection in Attention Mechanism
  (with Yingcong Li, Xuechen Zhang, and Samet Oymak)
  Neural Information Processing Systems (NeurIPS), 2023.
  Code Slide
  Spotlight (≈3% Acceptance)
• Fair Canonical Correlation Analysis
  (with Zhuoping Zhou, Bojian Hou, Boning Tong, Jia Xu, Yanbo Feng, Qi Long, and Li Shen)
  Neural Information Processing Systems (NeurIPS), 2023.
  Code Slide Poster
• Multi-Group Tensor Canonical Correlation Analysis
  (with Zhuoping Zhou, Boning Tong, Bojian Hou, Andrew J. Saykin, Qi Long, and Li Shen)
  The ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (ACM-BCB), 2023.
  Best Paper Award
• Fairness-Aware Class Imbalanced Learning on Multiple Subgroups
  (with Bojian Hou, Boning Tong, Qi Long, and Li Shen)
  In Uncertainty in Artificial Intelligence (UAI), pp. 2123-2133. PMLR, 2023.
  Code Poster
• Federated Multi-Sequence Stochastic Approximation with Local Hypergradient Estimation
  (with Mingchen Li, Pranay Sharma, and Samet Oymak)
  arXiv preprint, 2023.
  Code
• Fair Community Detection and Structure Learning in Heterogeneous Graphical Models
  (with Laura Balzano, and Alfred O. Hero)
  arXiv preprint, 2023.
  Slide
  

2022

• FedNest: Federated Bilevel, Minimax, and Compositional Optimization
  (with Mingchen Li, Christos Thrampoulidis, and Samet Oymak)
  International Conference on Machine Learning (ICML), 2022.
  Code Slide
  Oral (≈2% Acceptance)
• Data-Driven Control of Markov Jump Systems: Sample Complexity and Regret Bounds
  (with Zhe Du, Yahya Sattar, Laura Balzano, Necmiye Ozay, and Samet Oymak)
  American Control Conference (ACC), 2022.
• Certainty Equivalent Quadratic Control for Markov Jump Systems
  (with Zhe Du, Yahya Sattar, Laura Balzano, Samet Oymak, and Necmiye Ozay)
  American Control Conference (ACC), 2022.
• Truncated Matrix Completion - An Empirical Study
  (with Rishhabh Naik, Nisarg Trivedi, and Laura Balzano)
  In 30th European Signal Processing Conference (EUSIPCO), 2022.
• Dadam: A Consensus-Based Distributed Adaptive Gradient Method for Online Optimization
  (with Parvin Nazari, and George Michailidis)
  IEEE Transactions on Signal Processing 70, 6065-6079, 2022.
  Code
• Regularized and Smooth Double Core Tensor Factorization for Heterogeneous Data
  (with George Michailidis)
  Journal of Machine Learning Research 21, no.23, 2022.
  Code
• Grassmannian Optimization for Online Tensor Completion and Tracking with the t-SVD
  (with Kyle Gilman, and Laura Balzano)
  IEEE Transactions on Signal Processing 70, 2152-2167, 2022.
  Code

2021

• Identification and Adaptive Control of Markov Jump Systems: Sample Complexity and Regret Bounds
  (with Yahya Sattar, Zhe Du, Laura Balzano, Necmiye Ozay, and Samet Oymak)
  ICML Workshop on Reinforcement Learning Theory, 2021.
• Solving a Class of Nonconvex Min-Max Games Using Adaptive Momentum Methods
  (with Babak Barazandeh, George Michailidis)
  IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 3625-3629, 2021.
• Identification and Adaptive Control of Markov Jump Systems: Sample Complexity and Regret Bounds
  (with Yahya Sattar, Zhe Du, Laura Balzano, Necmiye Ozay, and Samet Oymak)
  arXiv preprint, 2021.

2020

• Adaptive First-and Zeroth-Order Methods for Weakly Convex Stochastic Optimization Problems
  (with Parvin Nazari, and George Michailidis)
  arXiv preprint, 2020.
• A Newton-Type Active Set Method for Nonlinear Optimization with Polyhedral Constraints
  (with William W. Hager)
  arXiv preprint, 2020.

2019

• Online Distributed Estimation of Principal Eigenspaces
  (with Mohamad Kazem Shirani Faradonbeh, and George Michailidis)
  IEEE Data Science Workshop (DSW), 2019.
• Adaptive Online Distributed Optimization in Dynamic Environments
  (with Parvin Nazari, and Esmaile Khorram)
  Optimization Methods and Software, 1-25, 2019.

2018

• Fast Randomized Algorithms for T-product Based Tensor Operations and Decompositions with Applications to Imaging Data
  (with George Michailidis)
  SIAM Journal on Imaging Sciences 11, no. 4: 2629-2664, 2018.
• Estimation of Graphical Models through Structured Norm Minimization
  (with George Michailidis)
  Journal of Machine Learning Research 18, no. 1, 2018.

2017 and Earlier

• A Nonmonotone PRP Conjugate Gradient Method for Solving Square and Under-Determined Systems of Equations
  (with Parvin Nazari, and Mohammad Reza Peyghami)
  Computers & Mathematics with Applications 73, no. 2: 339-354, 2017.
• A New Nonmonotone Adaptive Retrospective Trust Region Method for Unconstrained Optimization Problems
  (with Mohammad Reza Peyghami, and Fabian Bastin)
  Journal of Optimization Theory and Applications 167, no. 2: 676-692, 2015.
• A New Regularized Limited Memory BFGS-Type Method Based on Modified Secant Conditions for Unconstrained Optimization Problems
  (with Mohammad Reza Peyghami)
  Journal of Global Optimization 63, no. 4: 709-728, 2015.
• A Relaxed Nonmonotone Adaptive Trust Region Method for Solving Unconstrained Optimization Problems
  (with Mohammad Reza Peyghami)
  Computational Optimization and Applications 61, no. 2: 321-341, 2015.
  Code
• A New Trust Region Method for Solving Least-Square Transformation of System of Equalities and Inequalities
  (with Zeinab Saeidian, Mohammad Reza Peyghami, and Hamid Mesgarani)
  Optimization Letters 9, no. 2: 283-310, 2015.
• A New Nonmonotone Trust Region Method for Unconstrained Optimization Equipped by an Efficient Adaptive Radius
  (with Mohammad Reza Peyghami, and Hamid Mesgarani)
  Optimization Methods and Software 29, no. 4: 819-836, 2014.