| Sep 25, 2025 |  Our paper Distributionally Robust Optimization via Diffusion Ambiguity Modeling was accepted by OPT 2025 collocated with NeurIPS 2025! |
| Sep 20, 2025 | I was invited to give a talk at Hewlett Packard Enterprise Data Science Institute (HPE-DSI) on October. |
| Jul 15, 2025 | I will server in the Technical Program Committee of e-Energy’26 |
| Jul 1, 2025 | I will server in the Technical Program Committee of SIGMETRICS’26 |
| Jan 19, 2025 | Our paper Learning-Augmented Online Control for Decarbonizing Water Infrastructures has been accepted by e-Energy’25! This paper provides a learning-augmented control algorithm for a critical infrastructure: the municipal water supply systems. The algorithm guarantees the worst-case safety of water supply (e.g. for fire protection) while minizing the energy costs for pumping water. |
| Sep 25, 2024 | Our paper Online Budgeted Matching with General Bids has been accepted by Neurips 2024! It provides a provable algorithm to solve online budgeted matching problem with general bids. Manuscript will come out soon. |
| Sep 25, 2024 | Our paper Learning-Augmented Decentralized Online Convex Optimization in Networks has been accepted by Sigmetrics’25! It proposes a novel algorithm to provably robustify machine learning predictions for decentralized optimization in networks. |
| May 1, 2024 | Our paper Building Socially-Equitable Public Models was accepted by ICML 2024! Congratulations to Yejia and other co-authors! |
| Dec 13, 2023 | Our paper Online Allocation with Replenishable Budgets: Worst Case and Beyond was accepted by ACM SIGMETRICS 2024! Online resource allocation with replenishments is one of the key problems for carbon neutral and/or water self-sufficient computing and is important for network resource management, online advertising, etc. This paper provides fundamental algorithms with worst-case guarantees for this problem and proposes a learning-augmented algorithm to improve the statistical performance under worst-case guarantee. |
| Nov 28, 2023 | Our study on AI water Making AI Less “Thirsty”: Uncovering and Addressing the Secret Water Footprint of AI Models was featured by Nature Briefing! The paper quantitatively reveals the enormous water usage of AI, highlighting the necessity of addressing water footprint along with carbon footprint to enable truly sustainable AI. |
| Jun 21, 2023 | Check out our new paper on environmentally equitable AI Towards Environmentally Equitable AI via Geographical Load Balancing! This paper takes a first step toward addressing AI’s environmental inequity by optimization methods. |