Research

(alphabetical authorship marked with *)

Interactive demos of key ideas from my papers. Try them →

Journal Papers and Preprints

• High-Probability Polynomial-Time Complexity of Restarted Primal-Dual Hybrid Gradient Method for Linear Programming Zikai Xiong Mathematics of Operations Research, 2026 journal arXiv pdf demo
• Accessible Complexity Bounds for Restarted PDHG on Linear Programs with a Unique Optimizer Zikai Xiong Mathematics of Operations Research, 2026 journal arXiv pdf code slides demo ⭐ Winner, 2025 INFORMS Optimization Society Student Paper Prize
• Computational Guarantees for Restarted PDHG for LP based on "Limiting Error Ratios" and LP Sharpness Zikai Xiong and Robert M. Freund Mathematical Programming, 2026 journal arXiv pdf slides additional results ⭐ Honorable Mention, 2024 INFORMS Optimization Society Student Paper Prize
• Function-Free Optimization via Comparison Oracles Katya Scheinberg* and Zikai Xiong* Under revision arXiv pdf poster video demo animation
• Symmetry-dependence in Rounding of a Convex Body Zikai Xiong and Robert M. Freund Under review arXiv pdf demo
• The Role of Level-Set Geometry on the Performance of PDHG for Conic Linear Optimization Zikai Xiong and Robert M. Freund A subsequent version is under review arXiv pdf slides ⭐ Finalist, 2024 George Nicholson Student Paper Competition
• From an Interior Point to a Corner Point: Smart Crossover Dongdong Ge*, Chengwenjian Wang*, Zikai Xiong*, and Yinyu Ye* INFORMS Journal on Computing, 2025 journal arXiv pdf code slides
• A Practical GPU-Enhanced Matrix-Free Primal-Dual Method for Large-Scale Conic Programs Zhenwei Lin, Zikai Xiong, Dongdong Ge, and Yinyu Ye Under revision arXiv pdf code
• The Impact and Insights of Hidden City Ticketing: Analysis of a Choice-Based Network Revenue Management Model Jiacheng Guo*, Bo Jiang*, Zikai Xiong*, and Nanxi Zhang* Under revision for resubmission SSRN
• Using Taylor-Approximated Gradients to Improve the Frank-Wolfe Method for Empirical Risk Minimization Zikai Xiong and Robert M. Freund SIAM Journal on Optimization, 2024 journal arXiv pdf slides additional results
• Low-rank Traffic Matrix Completion with Marginal Information Zikai Xiong, Yimin Wei, Renjie Xu, and Yanwei Xu Journal of Computational and Applied Mathematics, 2022 journal pdf

Papers Published at Conferences

• Fair Wasserstein Coresets Zikai Xiong, Niccolò Dalmasso, Shubham Sharma, Freddy Lecue, Daniele Magazzeni, Vamsi K. Potluru, Tucker Balch, and Manuela Veloso Preliminary version appeared in NeurIPS 2023 Workshop SyntheticData4ML (Oral Presentation)
  Neural Information Processing Systems (NeurIPS 2024) proceedings arXiv pdf workshop version
• FairWASP: Fast and Optimal Fair Wasserstein Pre-processing Zikai Xiong, Niccolò Dalmasso, Alan Mishler, Vamsi Potluru, Tucker Balch, and Manuela Veloso AAAI Conference on Artificial Intelligence (AAAI 2024) proceedings arXiv pdf demo
• Learning from Multiple Annotator Noisy Labels via Sample-wise Label Fusion Zhengqi Gao, Fan-Keng Sun, Mingran Yang, Sucheng Ren, Zikai Xiong, et al. European Conference on Computer Vision (ECCV 2022) proceedings arXiv pdf code
• Interior-Point Methods Strike Back: Solving the Wasserstein Barycenter Problem Dongdong Ge*, Haoyue Wang*, Zikai Xiong*, and Yinyu Ye* Neural Information Processing Systems (NeurIPS 2019) proceedings arXiv pdf code slides

PhD Dissertation

• New Theory and New Practical Methods for Solving Large-Scale Linear and Conic Optimization PhD Dissertation DSpace@MIT pdf ⭐ Second Place, 2025 George B. Dantzig Dissertation Award

Technical Reports

• A Technical Note on the Implementation and Use of PDCS Zhenwei Lin, Zikai Xiong, Dongdong Ge, and Yinyu Ye arXiv pdf
• On the Relation Between LP Sharpness and Limiting Error Ratio and Complexity Implications for Restarted PDHG Zikai Xiong and Robert M. Freund arXiv pdf

Software

• PDCS A GPU-enhanced matrix-free primal-dual solver for large-scale conic programs, with JuMP and CVXPY interfaces. code
• Smart Crossover A Python package that finds optimal basic solutions from interior-point solutions. code
• PDLP Complexity Bound A Julia toolkit that computes the accessible iteration bounds of restarted PDHG for LP. code
• MAAIPM A MATLAB solver for exactly solving Wasserstein barycenter problems using interior-point methods. code

Patent Applications

• Method and System for Generating Fair Synthetic Representative Data via Optimal Transport Niccolò Dalmasso, Zikai Xiong, Vamsi Krishna Potluru, Tucker Richard Balch, and Manuela Veloso U.S. Patent Application Publication No. US 2025/0181999 A1, pending, 2025 Google Patents
• Method and System for Pre-processing Data for Algorithmic Fairness via Optimal Transport Niccolò Dalmasso, Zikai Xiong, Alan Mishler, Vamsi Krishna Potluru, Tucker Richard Balch, and Manuela Veloso U.S. Patent Application Publication No. US 2025/0173389 A1, pending, 2025 Google Patents