学术报告报告题目1：Sparse Decentralized Federated Learning
报告人：孔令臣 教授（北京交通大学）
报告时间：2025年9月27日（星期六） 9：15-9：50
报告摘要：Decentralized Federated Learning (DFL) enables collaborative model training without a central server but faces challenges in efficiency, stability, and trustworthiness due to communication and computational limitations among distributed nodes. To address these critical issues, we introduce a sparsity constraint on the shared model, leading to Sparse DFL (SDFL), and propose a novel algorithm, CEPS. The sparsity constraint facilitates the use of one-bit compressive sensing to transmit one-bit information between partially selected neighbour nodes at specific steps, thereby significantly improving communication efficiency. Moreover, we integrate differential privacy into the algorithm to ensure privacy preservation and bolster the trustworthiness of the learning process. Furthermore, CEPS is underpinned by theoretical guarantees regarding both convergence and privacy. Numerical experiments validate the effectiveness of the proposed algorithm in improving communication and computation efficiency while maintaining a high level of trustworthiness.
报告人简介：孔令臣，北京交通大学教授，博士生导师，中国运筹学会数学规划分会理事长，北京交通大学数学与统计学院副院长。主要从事对称锥互补问题和最优化、高维数据分析、统计优化与学习、医学成像等方面的研究。在《Mathematical Programming》、《SIAM Journal on Optimization》、《Statistica Sinica》、《IEEE Transactions on Pattern Analysis and Machine Intelligence》、《Technometrics》、《IEEE Transactions on Signal Processing》和《Electronic Journal of Statistics》等期刊发表论文60余篇。2005年获山东省高等教育教学成果奖三等奖，2012年获中国运筹学会青年奖，2018年获得北京市高等教育教学成果奖一等奖，2022年获教育部自然科学奖二等奖和北京市高等教育教学成果奖二等奖。
报告题目2：Non-convex Pose Graph Optimization in SLAM via Proximal Linearized Riemannian ADMM
报告人：韩德仁 教授（北京航空航天大学）
报告时间：2025年9月27日（星期六） 9：50-10：25
报告摘要：Pose graph optimization (PGO) is a well-known technique for solving the pose-based simultaneous localization and mapping (SLAM) problem. In this paper, we represent the rotation and translation by a unit quaternion and a three-dimensional vector, and propose a new PGO model based on the von Mises-Fisher distribution. The constraints derived from the unit quaternions are spherical manifolds, and the projectiononto the constraints can be calculated by normalization. Then a proximal linearized Riemannian alternating direction method of multipliers (PieADMM) is developed to solve the proposed model, which not only has low memory requirements, but also can update the poses in parallel. Furthermore, we establish the iteration complexity of 
of PieADMM for finding an ϵ-stationary solution of our model. The efficiency of our proposed algorithm is demonstrated by numerical experiments on two synthetic and four 3D SLAM benchmark datasets.
报告人简介：韩德仁，教授，博士生导师，北京航空航天大学数学科学学院院长、教育部数学类专业教指委秘书长。从事大规模优化、变分不等式问题及其应用研究工作，发表多篇学术论文。曾获中国运筹学会青年科技奖，江苏省科学技术奖等奖项;主持国家自然科学基金重点项目、杰出青年基金项目等多项项目。担任中国运筹学会副理事长、算法软件与应用分会理事长；《数值计算与计算机应用》、《Journal of the Operations Research Society of China》、《Journal of Global Optimization》、《Asia-Pacific Journal of Operational Research》编委。
报告题目3：经典投影一阶方法的新进展
报告人：夏勇 教授（北京航空航天大学）
报告时间：2025年9月27日（星期六） 10：45-11：20
报告摘要：本报告介绍我们近期在经典投影一阶方法方面的一些进展。第一部分重新审视经典的投影次梯度法，建立了经典消失步长的最优遍历收敛速度。此外我们还引入了弱遍历的概念，明晰教科书中所陷入的误区。第二部分进一步将经典投影次梯度法所依赖的函数值Lipschitz假设给予完全去除。第三部分介绍如何用投影梯度法全局求解经典的信赖域子问题这一非凸优化问题。
报告人简介：夏勇，北京航空航天大学教授，博士生导师，数学科学学院原副院长。2002年毕业于北京大学，2007年博士毕业于中国科学院，师从袁亚湘院士，2013年北京青年英才，2018年国家优青，在Math.Program.、SIAM J.Optim.、ICML、NeurIPS等期刊/会议发表论文100余篇。任中国运筹学会理事、中国运筹学会科普委员会副主任、数学规划分会常务理事、算法软件与应用分会常务理事、北京运筹学会常务理事、中国现场统计研究会贝叶斯统计分会常务理事，《Journal of the Operations Research Society of China》《Communications in Optimization Theory》期刊编委。获2023 MMOR最佳论文奖。代表性工作包括针对经典二次指派问题提出新模型，被国际国内同行命名为Xia-Yuan线性化，其线性规划松弛被称为Xia-Yuan界。
报告题目4：Gradient Norm Regularization Second-Order Algorithms for Solving
Nonconvex-Strongly Concave Minimax Problems
报告人：徐姿 教授（上海大学）
报告时间：2025年9月27日（星期六） 11：20-11：55
报告摘要：In this talk, we study second-order algorithms for the convex-concave minimax problem, which has attracted much attention in many fields such as machine learning in recent years. We propose a Lipschitz-free cubic regularization (LF-CR) algorithm for solving the convex-concave minimax optimization problem without knowing the Lipschitz constant. It can be shown that the iteration complexity of the LF-CR algorithm to obtain an $\epsilon$-optimal solution with respect to the restricted primal-dual gap is upper bounded by $\mathcal{O}(\rho^{2/3}\|z_0-z^*\|^2\epsilon^{-2/3})$, where $z_0=(x_0,y_0)$ is a pair of initial points, $z^*=(x^*,y^*)$ is a pair of optimal solutions, and $\rho$ is the Lipschitz constant. We further propose a fully parameter-free cubic regularization (FF-CR) algorithm that does not require any parameters of the problem, including the Lipschitz constant and the upper bound of the distance from the initial point to the optimal solution. We also prove that the iteration complexity of the FF-CR algorithm to obtain an $\epsilon$-optimal solution with respect to the gradient norm is upper bounded by $\mathcal{O}(\rho^{2/3}\|z_0-z^*\|^{4/3}\epsilon^{-2/3}) $. Numerical experiments show the efficiency of both algorithms. {\color{blue}To the best of our knowledge, the proposed FF-CR algorithm is a completely parameter-free second-order algorithm, and its iteration complexity is currently the best in terms of $\epsilon$ under the termination criterion of the gradient norm.
报告人简介：徐姿，上海大学理学院教授、博士生导师。主要研究方向是最优化理论与方法及在机器学习等领域中的应用，成果在Mathematical Programming，SIAM Journal on Optimization、Journal of Machine Learning Research、IEEE JSAC等国际著名期刊上发表论文 40余篇。主持国家自然科学基金项目4项和上海市自然基金项目1项。担任中国运筹学会数学规划分会常务理事、上海市运筹学会理事。现任Springer 旗下优化期刊J. Global Optim.客座编委（Guest Editor）；担任国际期刊 JORSC、PLOS One和Numerical Algebra, Control & Optimization编委。曾应邀赴美国明尼苏达大学、香港中文大学、香港理工大学等机构学术访问和交流。2020年获得中国运筹学会科学技术奖青年科技奖。2024年入选上海市东方英才计划拔尖项目。
报告题目5：Stratification for Nonlinear Semidefinite Programming
报告人：丁超 研究员（中国科学院数学与系统科学研究院）
报告时间：2025年9月27日（星期六） 14：15-14：50
报告摘要：Nonlinear semidefinite programs (NLSDPs) are traditionally analyzed through nonsmooth KKT systems in the ambient space, which necessitates strong regularity assumptions and obscures the reasons why Newton-type methods fail near degeneracy. In this talk, we introduce a stratification framework that resolves this tension by moving the analysis onto the strata of the symmetric matrices space. First, we prove that transversality for stratification is equivalent to W-SRCQ, establish its stability along the relevant strata, and show the genericity of W-SRCQ. Under this framework, the cone projection $\Pi_{\mathbb S_+^n}$ becomes a $C^\infty$ mapping between the appropriate strata and the ambient space; thus, the KKT system, nonsmooth globally but smooth on strata, can be solved by a strata Gauss–Newton method. We further prove that the restricted KKT mapping is a locally Lipschitz homeomorphism on the pertinent stratum if and only if W-SOC and W-SRCQ hold, yielding a local quadratic convergence rate for the strata Gauss–Newton method. Finally, we propose a globalized strategy tailored to the stratified geometry and establish global convergence, along with a quadratic local rate.
报告人简介：丁超，中国科学院数学与系统科学研究院应用数学研究所研究员，2012年于新加坡国立大学数学系毕业获得博士学位。研究方向为矩阵优化理论、算法及其应用以及大数据优化。围绕矩阵优化问题的理论、算法以及相关数据科学实际应用。丁超博士与国内外的合作者一起取得了一系列创新性研究成果，在包括《Mathematical Programming》、《SIAM Journal on Optimization》等数学优化权威期刊上发表多篇学术论文。目前担任中国运筹学会数学规划分会秘书长、亚太运筹学杂志《Asia-Pacific Journal of Operational Research》的编委。2016年获得中国运筹学会科学技术奖青年科技奖，一等奖。
报告题目6：A Single-Loop Algorithm for Decentralized Bilevel Optimization
报告人：杨俊锋 教授（南京大学）
报告时间：2025年9月27日（星期六） 14：50-15：25
报告摘要：Bilevel optimization (BO) has gained significant attention in recent years due to its broad applications in machine learning. In this talk, we focus on decentralized BO and proposes a novel single-loop algorithm for solving it with a strongly convex lower-level problem. Our approach is a fully single-loop method that approximates the hypergradient using only two matrix-vector multiplications per iteration. Our algorithm does not require any gradient heterogeneity assumption and achieves the best-known convergence rate for BO algorithms. We also present experimental results on hyperparameter optimization problems using both synthetic and MNIST datasets, which demonstrate the efficiency of our proposed algorithm.
报告人简介：杨俊锋，南京大学数学学院教授，博士生导师、副院长。2009年7月起在南京大学数学学院工作，主要从事最优化计算方法及其应用研究，在SIAM系列、Mathematics of Operations Research、Mathematics of Computation等杂志上发表论文40余篇，开发图像去模糊软代码包FTVd，压缩感知一模解码代码包YALL1，核磁共振图像复原代码包RecPF等。先后主持国家自然科学基金项目6项（国家优秀青年基金1项，面上项目3项，青年项目1项，天元访问学者项目1项）。获中国运筹学会青年科技奖、入选教育部新世纪优秀人才支持计划等，2020-2023年连续4年入选爱思唯尔中国高被引学者。担任中国运筹学会理事等，担任《计算数学》《ASVAO》《NACO》《SOIC》杂志编委、《Optimization in Engineering》客座编委等。
报告题目7：An Inexact Proximal Framework for Nonsmooth Riemannian Difference-of-Convex Optimization
报告人：姜波 教授（南京师范大学）
报告时间：2025年9月27日（星期六） 15：25-16：00
报告摘要：In this talk, we study nonsmooth Riemannian optimization models where the objective combines a smooth function with a nonsmooth difference-of-convex (DC) term, showing their equivalence to $\ell_0$-regularized or $\ell_0$-constrained problems. To solve them, we propose an inexact Riemannian proximal DC (iRPDC) framework with a novel inexactness criterion that enables efficient first-order subproblem solutions and an adaptive linesearch. The framework guarantees convergence to an $\epsilon$-Riemannian critical point in $\mathcal{O}(\epsilon^{-2})$ outer iterations, with one realization achieving the best-known overall complexity bound of $\mathcal{O}(\epsilon^{-3})$. Numerical results on sparse principal component analysis illustrate the flexibility of the DC formulation and the strong practical performance of our algorithms.
报告人简介：姜波，南京师范大学数学科学学院教授、博士生导师。2008 年本科毕业于中国石油大学（华东），2013 年博士毕业于中国科学院数学与系统科学研究院，2014 年 8 月起任职于南京师范大学。主要研究方向为流形约束优化算法与理论，在 Math. Program.、SIAM J. Optim.、SIAM J. Sci. Comput.、IEEE 汇刊以及 NeurIPS 等国际重要期刊和会议发表多篇学术论文。现主持国家自然科学基金青年科学基金项目（B 类）。曾入选第三届中国科协青年人才托举工程，获 2022 年中国运筹学会青年科技奖，并于 2024 年入选江苏省“333 工程”第三层次培养对象。现任中国运筹学会算法软件与应用分会理事、数学规划分会青年理事和江苏省运筹学会理事。
报告题目8：A Randomized Feasible Algorithm for Optimization with Orthogonal Constraints
报告人：范金燕 教授（上海交通大学）
报告时间：2025年9月27日（星期六） 16：20-16：55
报告摘要：In this talk, we present a randomized feasible algorithm for optimization over the Stiefel manifold, where only some randomly chosen columns of the variable matrix are updated at each iteration. It is proved that the sequence of Riemannian gradients generated by the algorithm converges to zero with probability one. Numerical results show that the algorithm is efficient, especially for the problems when the matrices involved are sparse.
报告人简介：范金燕，上海交通大学数学科学学院教授。主要从事最优化理论、方法及应用研究，在非线性最小二乘、无导数优化、张量计算等方面取得了一系列成果,主要发表在《Mathematical Programming》《SIAM Journal on Matrix Analysis and Applications》《Mathematics of Computation》等国际重要期刊。现担任中国运筹学会数学规划分会副理事长，《中国运筹学会会刊（Journal of the Operations Research Society of China）》、《计算数学》等杂志编委。曾获中国青年女科学家奖、中国青年科技奖，入选国家“万人计划”科技创新领军人才。
报告题目9：Cyclic stochastic gradient method
报告人：孙聪 教授（北京邮电大学）
报告时间：2025年9月27日（星期六） 16：55-17：30
报告摘要：The cyclic stepsize update strategy is proposed for stochastic gradient method. The stepsize is updated cyclicly, where the first two stepsizes use the approxiamted Cauchy steps and the rest apply the fixed stepsize. The step-ahead BB stepsize is used for the Cauchy step approximation. The method combines with both monotone and nonmonotone linesearches. The convergence properties are analyzed under different types of problems, where the theoretical results are proved without the interpolation condition assumption. The numerical tests show good performances of the proposed methods compared to other first order stochastic methods.
报告人简介：孙聪，北京邮电大学数学科学学院教授、博士生导师。2008年本科毕业于北京邮电大学理学院，2013年博士毕业于中国科学院数学与系统科学研究院。她的主要研究领域是非线性优化方法，特别是优化在信号处理中的应用。她曾获第三届中国科协青年托举人才工程的资助，入选北京邮电大学1551人才计划。孙聪博士发表论文三十余篇，其中包括IEEE Transactions on Signal Processing等信号处理领域顶级期刊和会议等，工作曾获北京运筹学会青年论文奖。她目前是中国运筹学会理事、副秘书长，中国运筹学会数学与智能分会理事，北京市运筹学会理事。
报告题目10：Dynamic Stochastic Approximation Jacobi-type ADMM Method for Two-stage Stochastic Generalized Nash Equilibrium Problems
报告人：孙海琳 教授（南京师范大学）
报告时间：2025年9月27日（星期六） 17：30-18：05
报告摘要：This paper studies a specific class of two-stage stochastic generalized Nash equilibrium problems (SGNEPs), where each player engages in a two-stage sequential decision-making process in a random environment: they make a decision in the current (first) stage and compete with one another, followed by making a decision in the future (second) stage. This type of two-stage SGNEPs is widely found in fields such as production and manufacturing, transportation logistics, and portfolio management. From the perspective of solving the problem, the main difference between two-stage SGNEPs and single-stage SGNEPs is the need to handle the optimal value function of the second-stage problem, which does not have an explicit expression. To overcome this difficulty, an accelerated primal-dual method (APDM) is proposed in the paper to obtain an $\epsilon$-subgradient of the second-stage optimal value function, achieving a convergence rate of $\mathcal{O}\left(\frac{1}{\sqrt{N}}\right)$. Using this $\epsilon$-subgradient along with a variance reduction technique, a dynamic stochastic approximation Jacobi-type Alternating Direction Method of Multipliers (DSA-JADMM) method is proposed and applied to solve two-stage SGNEPs. This algorithm represents an inexact stochastic version of the Jacobi-type ADMM, as it computes an $\epsilon$-subgradient for the second stage randomly at each iteration using APDM. It is also demonstrated that the algorithm can converge to a weak $\epsilon$-variational equilibrium point of two-stage SGNEPs with a convergence rate of $\mathcal{O}\left(\frac{1}{\sqrt{K}}\right)$. To validate the effectiveness of the DSA-JADMM, preliminary numerical experiments are conducted.
报告人简介：孙海琳博士是南京师范大学数学科学学院教授、副院长。他于2007年在吉林大学获得统计学学士学位，2013年毕业于哈尔滨工业大学，获数学博士学位。在其博士期间，他在英国南安普顿大学和香港理工大学联合培养。2015-2017年在香港理工大学应用数学系做博士后研究。2018年获中国运筹学会青年科技奖和江苏省数学成就奖，主持国家自然科学基金优秀青年科学基金项目、面上项目和青年科学基金项目。他的研究领域包括随机优化，分布鲁棒优化、随机变分不等式及其在投资组合、风险管理和经济学模型上的应用。他在包括《Mathematical Programming》、《SIAM Journal on Optimization》、《Mathematics of Operations Research》等国际权威期刊发表了二十多篇论文，担任《运筹学学报》、《计算数学》、《Journal of Optimization Theory and Applications》、《Asia-Pacific Journal of Operational Research》、《Numerical Algebra, Control and Optimization》等期刊编委。
报告地点：37号楼-3A02报告厅
邀请人：潘少华 教授、贲树军 副教授
欢迎广大师生前往！
华南理工大学数学学院
2025年9月24日