arXiv subjects

Seong-Hee Kim

Publications and source records attributed to Seong-Hee Kim.

6 recordsLinked to original sources

arXiv2026

Indifference-Zone Relaxation Procedures for Finding Feasible Systems

We consider the problem of finding feasible systems with respect to stochastic constraints when system performance is evaluated through simulation. Our objective is to solve this problem with high computational efficiency and statistical validity. Existing indifference-zone (IZ) procedures introduce a fixed tolerance level, which denotes how much deviation the decision-maker is willing to accept from the threshold in the constraint. These procedures are developed under the assumption that all systems' performance measures are exactly the tolerance level away from the threshold, leading to unnecessary simulations. In contrast, IZ-free procedures, which eliminate the tolerance level, perform well when systems' performance measures are far from the threshold. However, they may significantly underperform compared to IZ procedures when systems' performance measures are close to the threshold. To address these challenges, we propose the Indifference-Zone Relaxation (IZR) procedure, IZR introduces a set of relaxed tolerance levels and utilizes two subroutines for each level: one to identify systems that are clearly feasible and the other to exclude those that are clearly infeasible. We also develop the IZR procedure with estimation (IZE), which introduces two relaxed tolerance levels for each system and constraint: one matching the original tolerance level and the other based on an estimate of the system's performance measure. By employing different tolerance levels, these procedures facilitate early feasibility determination with statistical validity. We prove that IZR and IZE determine system feasibility with the desired probability and show through experiments that they significantly reduce the number of observations required compared to an existing procedure.

eess.SY↗

arXiv2026

Feasibility Determination for Subjective Probability Constraints

We consider the problem of determining feasible systems from a finite set of simulated alternatives with respect to probability constraints, where the observations from stochastic simulations are Bernoulli distributed. Most statistically valid procedures for feasibility determination focus on constraints on the means of normally distributed observations. Although these procedures can be adapted to Bernoulli-distributed data by treating batch means as basic observations, achieving approximate normality often requires a large batch size, potentially leading to the unnecessary waste of observations in reaching a decision. This paper proposes a procedure that utilizes the Bernoulli-distributed observations directly to determine feasibility. In addition, we incorporate subjective constraints, allowing for multiple thresholds for each constraint. We demonstrate that our proposed procedure is statistically valid and that it outperforms an existing feasibility determination procedure for subjective constraints originally developed for normally distributed observations. Furthermore, we propose two heuristic feasibility check approaches for thresholds that are sequentially added by decision makers, allowing thresholds to be tightened when many systems are feasible or relaxed when no feasible system exists. We show by experiments that the proposed procedures can efficiently provide feasibility decisions to systems with respect to all thresholds considered.

math.OC↗

arXiv2025

Distribution-Free Online Change Detection for Low-Rank Images

We present a distribution-free CUSUM procedure designed for online change detection in a time series of low-rank images, particularly when the change causes a mean shift. We represent images as matrix data and allow for temporal dependence, in addition to inherent spatial dependence, before and after the change. The marginal distributions are assumed to be general, not limited to any specific parametric distribution. We propose new monitoring statistics that utilize the low-rank structure of the in-control mean matrix. Additionally, we study the properties of the proposed detection procedure, assessing whether the monitoring statistics effectively capture a mean shift and evaluating the rate of increase in the average run length relative to the control limit in both the in-control and out-of-control cases. The effectiveness of our procedure is demonstrated through simulated and real data experiments.

stat.ME↗

arXiv2025

Fixed Confidence and Fixed Tolerance Bi-level Optimization for Selecting the Best Optimized System

In this paper, we study a fixed-confidence, fixed-tolerance formulation of a class of stochastic bi-level optimization problems, where the upper-level problem selects from a finite set of systems based on a performance metric, and the lower-level problem optimizes continuous decision variables for each system. Notably, the objective functions for the upper and lower levels can differ. This class of problems has a wide range of applications, including model selection, ranking and selection under input uncertainty, and optimal design. To address this, we propose a multi-stage Pruning-Optimization framework that alternates between comparing the performance of different systems (Pruning) and optimizing systems (Optimization). % In the Pruning stage, we design a sequential algorithm that identifies and eliminates inferior systems through systematic performance evaluations. In the Optimization stage, the goal is to solve for a near-optimal solution that meets specified confidence and tolerance requirements. This multi-stage framework is designed to enhance computational efficiency by pruning inferior systems with high tolerance early on, thereby avoiding unnecessary computational efforts. We demonstrate the effectiveness of the proposed algorithm through both theoretical analysis of statistical validity and sample complexity and numerical experiments.

math.OC↗

arXiv2021

Efficient Fully Sequential Indifference-Zone Procedures Using Properties of Multidimensional Brownian Motion Exiting a Sphere

We consider a ranking and selection (R&S) problem with the goal to select a system with the largest or smallest expected performance measure among a number of simulated systems with a pre-specified probability of correct selection. Fully sequential procedures take one observation from each survived system and eliminate inferior systems when there is clear statistical evidence that they are inferior. Most fully sequential procedures make elimination decisions based on sample performances of each possible pair of survived systems and exploit the bound crossing properties of a univariate Brownian motion. In this paper, we present new fully sequential procedures with elimination decisions that are based on sample performances of all competing systems. Using properties of a multidimensional Brownian motion exiting a sphere, we derive heuristics that aim to achieve a given target probability of correct selection. We show that in practice the new procedures significantly outperform a widely used fully sequential procedure. Compared to BIZ, a recent fully-sequential procedure that uses statistics inspired by Bayes posterior probabilities, our procedures have better performance under difficult mean or variance configurations but similar performance under easy mean configurations.

stat.ME↗

arXiv2018

$\textsf{S}^3T$: An Efficient Score-Statistic for Spatio-Temporal Surveillance

We present an efficient score statistic, called the $\textsf{S}^3 \textsf{T}$ statistic, to detect the emergence of a spatially and temporally correlated signal from either fixed-sample or sequential data. The signal may cause a men shift and/or a change in the covariance structure. The score statistic can capture both spatial and temporal structures of the change and hence is particularly powerful in detecting weak signals. The score statistic is computationally efficient and statistically powerful. Our main theoretical contribution are accurate analytical approximations on the false alarm rate of the detection procedures, which can be used to calibrate the threshold analytically. Numerical experiments on simulated and real data demonstrate the good performance of our procedure for solar flame detection and water quality monitoring.

math.ST↗