We study the Heilbronn triangle problem, which involves placing \(n\) points in the unit square such that the minimum area of any triangle formed by these points is maximized. A straightforward maximin formulation of this problem is highly non-linear and non-convex due to the existence of bilinear terms and absolute value equations. We propose two … Read more
Distributionally robust optimization (DRO) has emerged as a powerful paradigm for reliable decision-making under uncertainty. This paper focuses on DRO with ambiguity sets defined via the Sinkhorn discrepancy: an entropy-regularized Wasserstein distance, referred to as Sinkhorn DRO. Existing work primarily addresses Sinkhorn DRO from a dual perspective, leveraging its formulation as a conditional stochastic optimization … Read more
We consider the design of smoothings of the (coordinate-wise) max function in $\mathbb{R}^d$ in the infinity norm. The LogSumExp function $f(x)=\ln(\sum^d_i\exp(x_i))$ provides a classical smoothing, differing from the max function in value by at most $\ln(d)$. We provide an elementary construction of a lower bound, establishing that every overestimating smoothing of the max function must … Read more
We consider a two-stage optimization problem with sparsity constraints, motivated by a common challenge in packaging logistics: minimizing the volume of transported air by optimizing the size and number of available packaging boxes, given the demand for order items. In the first stage, we select the optimal dimensions of the boxes, while in the second … Read more
We develop a unified framework for robust nonsmooth optimization problems with equilibrium constraints (UNMPEC). As a foundation, we study a robust nonsmooth nonlinear program with uncertainty in both the objective function and the inequality constraints (UNP). Using Clarke subdifferentials, we establish Karush–Kuhn–Tucker (KKT)–type necessary optimality conditions under an extended no–nonzero–abnormal–multiplier constraint qualification (ENNAMCQ). When the … Read more
In this paper, we revisit a classical adaptive stepsize strategy for gradient descent: the Polyak stepsize (PolyakGD), originally proposed in Polyak (1969). We study the convergence behavior of PolyakGD from two perspectives: tight worst-case analysis and universality across function classes. As our first main result, we establish the tightness of the known convergence rates of … Read more
We develop and study a subsampled cubic regularization method for finite-sum composite optimization problems, in which the function, gradient, and Hessian are estimated using possibly different sample sizes. By allowing each quantity to have its own sampling strategy, the proposed method offers greater flexibility to control the accuracy of the model components and to better … Read more
Transportation network design, or the problem of optimizing infrastructure for a societal goal, subject to individual travelers optimizing their behavior for their own preferences arises frequently in many contexts. However, it is also an NP-hard problem due to the leader-follower or bi-level structure involving a follower objective that is different from yet significantly affects the … Read more
Aligning large language models (LLMs) with human preferences has evolved from single-objective reward maximization to sophisticated multi-objective optimization. Real-world deployment requires balancing competing objectiveshelpfulness, harmlessness, honesty, instruction-following, and task-specic capabilitiesthat often conict. This survey provides a systematic taxonomy of multi-objective alignment techniques, organizing the rapidly growing literature into four categories: (1) Reward Decomposition approaches that … Read more
Modern supply chains face mounting uncertainty and scale, motivating the integration of Artificial Intelligence (AI) and Machine Learning (ML) with mathematical optimization to enable robust and adaptive decisions. We present a systematic review of 199 articles on tangible supply chains, categorizing how ML is used—primarily for parameter estimation and for solution generation—and proposing a taxonomy … Read more