邓琪，现任上海交通大学安泰经济与管理学院副教授（@浩然大厦815），此前在上海财经大学信息管理与工程学院工作，分别担任助理教授和副教授。分别于美国佛罗里达大学和上海交通大学获得计算机专业博士和学士学位。研究领域包括数学规划和机器学习，研究内容包括大规模优化算法，近期对AI和优化结合、端到端学习很感兴趣，近年来的研究成果发表在 Math Program、IJOC、MOR、POMS、 PNAS、 NeurIPS、ICML 等运筹优化与机器学习领域的期刊和会议上。获中国运筹学会青年科技奖（2026），目前主持国家自科与上海市面上项目，参与国家自科重大项目。

联系方式： qdeng24(at)sjtu.edu.cn (请手动将(at)替换为@)

招生信息

研究生：欢迎有志于从事学术研究工作的考生联系报考，要求自律、态度端正、有一定的英语写作基础、编程和数理能力较强。常年有硕士名额；博士名额不定。

目前在推进几个研究课题：

• LLM 优化方法：结合大模型结构与训练特性，设计优化算法及其高效实现；偏工程实现与动手能力（数学 30%，工程 70%）

• AI 与优化的双向结合：求解器嵌入神经网络的端到端学习（优化 → AI），以及 AI 加速的优化算法设计（AI → 优化）；理论与应用并重，关注金融和管理场景落地（数学 50%，工程 50%）

• 数学优化基础理论：研究不同问题设定下连续优化算法的理论性质与复杂度分析；以理论分析为主（数学 70%，工程 30%）

研究兴趣：连续优化的算法与复杂度理论；联合预测与优化的集成模型；能源与电力系统中的应用

近期主要工作：

Han, Q., Lin, Z., Liu, H., Chen, C., Deng, Q., Ge, D., & Ye, Y. (2026). Large-scale semidefinite programming with graphics processing units. Proceedings of the National Academy of Sciences, 123(40).

Linghu, Y., Liu, Z., & Deng, Q. (2026). A penalty approach for differentiation through black-box quadratic programming solvers. International Conference on Machine Learning (ICML 2026).

Deng, Q., Lan, G., & Lin, Z. (2026). Uniformly optimal and parameter-free first-order methods for convex and function-constrained optimization. INFORMS Journal on Computing.

Boob, D., Deng, Q., & Khalafi, M. (2026). First-order methods for stochastic variational inequality problems with function constraints. Mathematics of Operations Research.

Ren, Y., Xu, H., & Deng, Q. (2025). Accelerated distance-adaptive methods for Hölder smooth and convex optimization. Advances in Neural Information Processing Systems (NeurIPS), 38.

Han, Q., Li, C., Lin, Z., Chen, C., Deng, Q., Ge, D., Liu, H., & Ye, Y. (2025). A low-rank ADMM splitting approach for semidefinite programming. INFORMS Journal on Computing.

Boob, D., Deng, Q., & Lan, G. (2025). Level constrained first order methods for function constrained optimization. Mathematical Programming, 209(1), 1–61.

Lin, Z., & Deng, Q. (2024). Faster accelerated first-order methods for convex optimization with strongly convex function constraints. Advances in Neural Information Processing Systems (NeurIPS), 37.

Deng, Q., Feng, Q., Gao, W., Ge, D., Jiang, B., Jiang, Y., Liu, J., Liu, T., Xue, C., & Ye, Y. (2024). An enhanced ADMM-based interior point method for linear and conic optimization. INFORMS Journal on Computing.

Lin, Z., Xue, C., Deng, Q., & Ye, Y. (2024). A single-loop robust policy gradient method for robust Markov decision processes. International Conference on Machine Learning (ICML 2024).

Gao, W., & Deng, Q. (2024). Stochastic weakly convex optimization beyond Lipschitz continuity. International Conference on Machine Learning (ICML 2024).

Tan, J., Xue, C., Zhang, C., Deng, Q., Ge, D., & Ye, Y. (2024). A homogenization approach for gradient-dominated stochastic optimization. Conference on Uncertainty in Artificial Intelligence (UAI 2024).

Xie, C., Li, C., Zhang, C., Deng, Q., Ge, D., & Ye, Y. (2024). Trust region methods for nonconvex stochastic optimization beyond Lipschitz smoothness. Proceedings of the AAAI Conference on Artificial Intelligence.

Liu, J., Xie, C., Deng, Q., Ge, D., & Ye, Y. (2024). Sketched Newton value iteration for large-scale Markov decision processes. Proceedings of the AAAI Conference on Artificial Intelligence.

Lin, Z., Xia, J., Deng, Q., & Luo, L. (2024). Decentralized gradient-free methods for stochastic non-smooth non-convex optimization. Proceedings of the AAAI Conference on Artificial Intelligence.

王新迎, 闫冬, 施展, 张东霞, 邓琪, 林振炜 (2024). 机器学习赋能的优化算法及其在新型电力系统中的应用与展望. 中国电机工程学报, 44(16), 6367–6385.

Gao, W., & Deng, Q. (2023). Delayed algorithms for distributed stochastic weakly convex optimization. Advances in Neural Information Processing Systems (NeurIPS), 36.

Shi, Z., Wang, X., Yan, D., Chen, S., Lin, Z., Xia, J., & Deng, Q. (2023). An accelerated primal-dual method for semi-definite programming relaxation of optimal power flow. IET Energy Systems Integration, 5(4), 477–490.

Boob, D., Deng, Q., & Lan, G. (2023). Stochastic first-order methods for convex and nonconvex functional constrained optimization. Mathematical Programming, 197(1), 215–279.

Deng, Q., & Gao, W. (2021). Minibatch and momentum model-based methods for stochastic weakly convex optimization. Advances in Neural Information Processing Systems (NeurIPS), 34, 23115–23127.

邓琪, 高建军, 葛冬冬, 何斯迈, 江波, 李晓澄, 王子卓, 杨超林, 叶荫宇 (2020). 现代优化理论与应用. 中国科学：数学, 50(7), 899–968.

Deng, Q., & Lan, C. (2020). Efficiency of coordinate descent methods for structured nonconvex optimization. Joint European Conference on Machine Learning and Knowledge Discovery in Databases (ECML PKDD).

Boob, D., Deng, Q., Lan, G., & Wang, Y. (2020). A feasible level proximal point method for nonconvex sparse constrained optimization. Advances in Neural Information Processing Systems (NeurIPS), 33, 16773–16784.

Tan, Y., Paul, A. A., Deng, Q., & Wei, L. (2017). Mitigating inventory overstocking: Optimal order-up-to level to achieve a target fill rate over a finite horizon. Production and Operations Management, 26(11), 1971–1988.

数据结构