arXiv:2407. 04900v2 Announce Type: replace Abstract: Numerous existing studies have examined the performance of Sample Average Approximation (SAA) in the fundamental newsvendor problem.
arXiv:2609. 28263v1 Announce Type: new Abstract: The growth of large language model (LLM) inference and search services increases the scale of online linear programming problems, motivating computationally efficient algorithms.
The paper introduces SELECT, an algorithmic framework for satisficing regret minimization in bandit problems, achieving constant expected satisficing regret when a satisficing arm exists. A variant, SELECT‑LITE, further ensures a light‑tailed satisficing regret distribution while maintaining constant expected regret in the realizable case and sub‑linear standard regret otherwise. Experiments on synthetic data and a real‑world dynamic pricing scenario demonstrate the practical effectiveness of both algorithms.
arXiv:2607. 15229v1 Announce Type: new Abstract: We develop data-driven algorithms for maintaining $N$ independent identical machines under a \textit{block replacement policy}, in which each machine is replaced upon failure and all machines are jointly replaced at regular intervals of length $k$.
arXiv:2607. 10207v1 Announce Type: cross Abstract: Data-driven optimization often requires collecting data to estimate uncertain model parameters before solving the underlying decision problem.
arXiv:2606. 15600v1 Announce Type: cross Abstract: Cardinality-estimation (CE) research ranks estimators by q-error, yet it is well known that q-error is an imperfect proxy for query-plan quality.
The paper presents a data‑driven framework for multi‑period lost‑sales inventory control when demand is censored, meaning stockouts only reveal that demand exceeded the stocking level. It introduces a new cost decomposition for base‑stock policies and a biased sample‑average approximation (SAA) method, leading to two algorithms: an upper‑biased SAA that achieves near‑optimal sample complexity under an offline coverage condition, and a lower‑biased SAA that actively generates coverage to achieve near‑optimal online regret. The biased SAA approach offers a general principle for applying pessimism and optimism in settings with censored feedback.
arXiv:2608. 25551v1 Announce Type: new Abstract: Stochastic gradient descent (SGD) is typically analyzed at a deterministic horizon chosen before the algorithm is run, even though practical stopping decisions are made adaptively by inspecting the evolving trajectory.
The paper investigates stochastic contextual bandits in an agnostic setting, aiming to compete with the best policy in a given class without assuming realizability or specific loss/reward models. It introduces an algorithm that updates the policy each round by minimizing a pessimistic objective— a clipped inverse‑propensity estimate of the policy value plus a variance penalty— and proves the first fast regret rates relative to the best‑in‑class policy. By exploiting entropy assumptions on the policy class and a H"olderian error‑bound condition, the authors achieve fast best‑in‑class regret rates, including polylogarithmic rates in the parametric case, using a sequential self‑normalized maximal inequality for bounded martingale empirical processes to derive uniform variance‑adaptive confidence bounds and ensure pessimism under adaptive data collection.
arXiv:2607. 29460v1 Announce Type: new Abstract: Heavy-tailed distributions arise naturally in sequential decision-making problems such as financial investment, online advertising, and network management, where rare but extreme outcomes can dominate performance.
arXiv:2606. 14679v1 Announce Type: new Abstract: Online inventory optimization (OIO) is online convex optimization with physical memory: inventory carryover makes the feasible action set depend on the past.
arXiv:2312. 15427v3 Announce Type: replace Abstract: Stochastic optimization is a widely used approach for optimization under uncertainty, where uncertain input parameters are modeled by random variables.
The paper introduces a new approach to safety in contextual bandits with continuous actions, focusing on high‑probability constraints on the realized cost rather than expected cost. It presents the High‑Probability Constrained UCB algorithm, which balances reward exploration with conservative safety estimation, and provides theoretical regret guarantees for linear models and extensions to general function classes. Experiments demonstrate that this realized‑cost safety framework significantly reduces safety violations compared to expected‑cost constrained methods.