arXiv subjects

Yongho Shin

Publications and source records attributed to Yongho Shin.

14 recordsLinked to original sources

arXiv2026

Parsimonious Learning-Augmented Online Metric Matching

Learning-augmented algorithms have received significant attention in recent years, particularly in the context of online optimization. Motivated by the high computational cost of generating predictions, a growing line of work studies the tradeoff between performance guarantees and the number of predictions used in learning-augmented algorithms for problems such as caching and metrical task systems. In this paper, we extend this line of research to online metric matching by developing parsimonious learning-augmented algorithms and establishing lower bounds on their performance. Our approach extends the Follow-the-Prediction framework to the parsimonious setting by filling in a virtual prediction in the absence of an actual prediction, using an online metric matching algorithm that maintains good intermediate matchings throughout its execution. We complement our theoretical results with an empirical evaluation, demonstrating the practical effectiveness of our approach.

cs.DS↗

arXiv2026

Optimal Learning-Augmented Algorithm for Online Bidding

Recent advances in machine learning have spurred significant interest in learning-augmented algorithms, particularly for online optimization. A growing body of work has studied online bidding in this framework, aiming to characterize the trade-off between robustness and consistency. While this trade-off is fully understood for deterministic algorithms, a gap between upper and lower bounds remains in the randomized setting. In this paper, we close this gap by presenting a Pareto-optimal randomized learning-augmented algorithm for this problem. Our approach introduces the notion of a bidding profile, a novel framework for representing the distribution over bids generated by an algorithm. We show that any bidding algorithm can be reduced, without loss of generality, to one driven by a bidding profile, and we characterize the optimal profile via a system of delayed differential equations. Finally, we demonstrate the broader applicability of our approach by extending it to the linear search problem, yielding a significant improvement over prior learning-augmented algorithms for linear search.

cs.DS↗

arXiv2026

Servicing Matched Client Pairs with Facilities

We study Facility Location with Matching, a Facility Location problem where, given additional information about which pair of clients is compatible to be matched, we need to match as many clients as possible and assign each matched client pair to a same open facility at minimum total cost. The problem is motivated by match-making services relevant in, for example, video games or social apps. It naturally generalizes two prominent combinatorial optimization problems -- Uncapacitated Facility Location and Minimum-cost Maximum Matching. Facility Location with Matching also generalizes the Even-constrained Facility Location problem studied by Kim, Shin, and An (Algorithmica 2023). We propose a linear programming (LP) relaxation for this problem, and present a 3.868-approximation algorithm. Our algorithm leverages the work on bifactor-approximation algorithms (Byrka and Aardal, SICOMP 2012); our main technical contribution is a rerouting subroutine that reroutes a fractional solution to be supported on a fixed maximum matching with only small additional cost. For a special case where all clients are matched, we provide a refined algorithm achieving an approximation ratio of 2.218. As our algorithms are based on rounding an optimal solution to the LP relaxation, these approximation results also give the same upper bounds on the integrality gap of the relaxation.

cs.DS↗

arXiv2025

Post-Training and Test-Time Scaling of Generative Agent Behavior Models for Interactive Autonomous Driving

Learning interactive motion behaviors among multiple agents is a core challenge in autonomous driving. While imitation learning models generate realistic trajectories, they often inherit biases from datasets dominated by safe demonstrations, limiting robustness in safety-critical cases. Moreover, most studies rely on open-loop evaluation, overlooking compounding errors in closed-loop execution. We address these limitations with two complementary strategies. First, we propose Group Relative Behavior Optimization (GRBO), a reinforcement learning post-training method that fine-tunes pretrained behavior models via group relative advantage maximization with human regularization. Using only 10% of the training dataset, GRBO improves safety performance by over 40% while preserving behavioral realism. Second, we introduce Warm-K, a warm-started Top-K sampling strategy that balances consistency and diversity in motion selection. Our Warm-K method-based test-time scaling enhances behavioral consistency and reactivity at test time without retraining, mitigating covariate shift and reducing performance discrepancies. Demo videos are available in the supplementary material.

cs.RO↗

arXiv2025

Approximating (Weighted) Chromatic Correlation Clustering via Cluster LP

Correlation Clustering is a fundamental clustering problem that is generalized to Chromatic Correlation Clustering to incorporate categorical data. Both problems have been intensively studied, and recently, substantial improvements were obtained in the approximation algorithms for Correlation Clustering. At the heart of this success lies a new linear program (LP) formulation called the cluster LP; a natural question was whether this LP can be extended to Chromatic Correlation Clustering to enable similar success. We answer this question in the affirmative by presenting a $(2+\varepsilon)$-approximation algorithm for the problem using a chromatic cluster LP. We then consider Weighted Chromatic Correlation Clustering, in which edges have fractional weights satisfying the probability constraints, to show that our algorithm extends to this weighted version to yield the same approximation guarantee.

cs.DS↗

arXiv2025

Online Rounding for Set Cover under Subset Arrivals

A rounding scheme for set cover has served as an important component in design of approximation algorithms for the problem, and there exists an H_s-approximate rounding scheme, where s denotes the maximum subset size, directly implying an approximation algorithm with the same approximation guarantee. A rounding scheme has also been considered under some online models, and in particular, under the element arrival model used as a crucial subroutine in algorithms for online set cover, an O(log s)-competitive rounding scheme is known [Buchbinder, Chen, and Naor, SODA 2014]. On the other hand, under a more general model, called the subset arrival model, only a simple O(log n)-competitive rounding scheme is known, where n denotes the number of elements in the ground set. In this paper, we present an O(log^2 s)-competitive rounding scheme under the subset arrival model, with one mild assumption that s is known upfront. Using our rounding scheme, we immediately obtain an O(log^2 s)-approximation algorithm for multi-stage stochastic set cover, improving upon the existing algorithms [Swamy and Shmoys, SICOMP 2012; Byrka and Srinivasan, SIDMA 2018] when s is small enough compared to the number of stages and the number of elements. Lastly, for set cover with s = 2, also known as edge cover, we present a 1.8-competitive rounding scheme under the edge arrival model.

cs.DS↗

arXiv2025

Learning-Augmented Online Bipartite Fractional Matching

Online bipartite matching is a fundamental problem in online optimization, extensively studied both in its integral and fractional forms due to its theoretical significance and practical applications, such as online advertising and resource allocation. Motivated by recent progress in learning-augmented algorithms, we study online bipartite fractional matching when the algorithm is given advice in the form of a suggested matching in each iteration. We develop algorithms for both the vertex-weighted and unweighted variants that provably dominate the naive "coin flip" strategy of randomly choosing between the advice-following and advice-free algorithms. Moreover, our algorithm for the vertex-weighted setting extends to the AdWords problem under the small bids assumption, yielding a significant improvement over the seminal work of Mahdian, Nazerzadeh, and Saberi (EC 2007, TALG 2012). Complementing our positive results, we establish a hardness bound on the robustness-consistency tradeoff that is attainable by any algorithm. We empirically validate our algorithms through experiments on synthetic and real-world data.

cs.DS↗

arXiv2025

Improved Algorithms for Overlapping and Robust Clustering of Edge-Colored Hypergraphs: An LP-Based Combinatorial Approach

Clustering is a fundamental task in both machine learning and data mining. Among various methods, edge-colored clustering (ECC) has emerged as a useful approach for handling categorical data. Given a hypergraph with (hyper)edges labeled by colors, ECC aims to assign vertex colors to minimize the number of edges where the vertex color differs from the edge's color. However, traditional ECC has inherent limitations, as it enforces a nonoverlapping and exhaustive clustering. To tackle these limitations, three versions of ECC have been studied: Local ECC and Global ECC, which allow overlapping clusters, and Robust ECC, which accounts for vertex outliers. For these problems, both linear programming (LP) rounding algorithms and greedy combinatorial algorithms have been proposed. While these LP-rounding algorithms provide high-quality solutions, they demand substantial computation time; the greedy algorithms, on the other hand, run very fast but often compromise solution quality. In this paper, we present an algorithmic framework that combines the strengths of LP with the computational efficiency of combinatorial algorithms. Both experimental and theoretical analyses show that our algorithms efficiently produce high-quality solutions for all three problems: Local, Global, and Robust ECC. We complement our algorithmic contributions with complexity-theoretic inapproximability results and integrality gap bounds, which suggest that significant theoretical improvements are unlikely. Our results also answer two open questions previously raised in the literature.

cs.LG↗

arXiv2024

Textual Training for the Hassle-Free Removal of Unwanted Visual Data: Case Studies on OOD and Hateful Image Detection

In our study, we explore methods for detecting unwanted content lurking in visual datasets. We provide a theoretical analysis demonstrating that a model capable of successfully partitioning visual data can be obtained using only textual data. Based on the analysis, we propose Hassle-Free Textual Training (HFTT), a streamlined method capable of acquiring detectors for unwanted visual content, using only synthetic textual data in conjunction with pre-trained vision-language models. HFTT features an innovative objective function that significantly reduces the necessity for human involvement in data annotation. Furthermore, HFTT employs a clever textual data synthesis method, effectively emulating the integration of unknown visual data distribution into the training process at no extra cost. The unique characteristics of HFTT extend its utility beyond traditional out-of-distribution detection, making it applicable to tasks that address more abstract concepts. We complement our analyses with experiments in out-of-distribution detection and hateful image detection. Our codes are available at https://github.com/Saehyung-Lee/HFTT

cs.CV↗

arXiv2023

On Optimal Consistency-Robustness Trade-Off for Learning-Augmented Multi-Option Ski Rental

The learning-augmented multi-option ski rental problem generalizes the classical ski rental problem in two ways: the algorithm is provided with a prediction on the number of days we can ski, and the ski rental options now come with a variety of rental periods and prices to choose from, unlike the classical two-option setting. Subsequent to the initial study of the multi-option ski rental problem (without learning augmentation) due to Zhang, Poon, and Xu, significant progress has been made for this problem recently in particular. The problem is very well understood when we relinquish one of the two generalizations -- for the learning-augmented classical ski rental problem, algorithms giving best-possible trade-off between consistency and robustness exist; for the multi-option ski rental problem without learning augmentation, deterministic/randomized algorithms giving the best-possible competitiveness have been found. However, in presence of both generalizations, there remained a huge gap between the algorithmic and impossibility results. In fact, for randomized algorithms, we did not have any nontrivial lower bounds on the consistency-robustness trade-off before. This paper bridges this gap for both deterministic and randomized algorithms. For deterministic algorithms, we present a best-possible algorithm that completely matches the known lower bound. For randomized algorithms, we show the first nontrivial lower bound on the consistency-robustness trade-off, and also present an improved randomized algorithm. Our algorithm matches our lower bound on robustness within a factor of e/2 when the consistency is at most 1.086.

cs.DS↗

arXiv2023

Improved Learning-Augmented Algorithms for the Multi-Option Ski Rental Problem via Best-Possible Competitive Analysis

In this paper, we present improved learning-augmented algorithms for the multi-option ski rental problem. Learning-augmented algorithms take ML predictions as an added part of the input and incorporates these predictions in solving the given problem. Due to their unique strength that combines the power of ML predictions with rigorous performance guarantees, they have been extensively studied in the context of online optimization problems. Even though ski rental problems are one of the canonical problems in the field of online optimization, only deterministic algorithms were previously known for multi-option ski rental, with or without learning augmentation. We present the first randomized learning-augmented algorithm for this problem, surpassing previous performance guarantees given by deterministic algorithms. Our learning-augmented algorithm is based on a new, provably best-possible randomized competitive algorithm for the problem. Our results are further complemented by lower bounds for deterministic and randomized algorithms, and computational experiments evaluating our algorithms' performance improvements.

cs.DS↗

arXiv2021

Making Three Out of Two: Three-Way Online Correlated Selection

Two-way online correlated selection (two-way OCS) is an online algorithm that, at each timestep, takes a pair of elements from the ground set and irrevocably chooses one of the two elements, while ensuring negative correlation in the algorithm's choices. Whilst OCS was initially invented by Fahrbach, Huang, Tao, and Zadimoghaddam to solve the edge-weighted online bipartite matching problem, it is an interesting technique on its own due to its capability of introducing a powerful algorithmic tool, namely negative correlation, to online algorithms. As such, Fahrbach et al. posed two tantalizing open questions in their paper, one of which was the following: Can we obtain n-way OCS for n>2, in which the algorithm can be given n>2 elements to choose from at each timestep? In this paper, we affirmatively answer this open question by presenting a three-way OCS. Our algorithm uses two-way OCS as its building block and is simple to describe; however, as it internally runs two instances of two-way OCS, one of which is fed with the output of the other, the final output probability distribution becomes highly elusive. We tackle this difficulty by approximating the output distribution of OCS by a flat, less correlated function and using it as a safe "surrogate" of the real distribution. Our three-way OCS also yields a 0.5093-competitive algorithm for edge-weighted online matching, demonstrating its usefulness.

cs.DS↗

arXiv2020

Online Graph Matching Problems with a Worst-Case Reassignment Budget

In the online bipartite matching with reassignments problem, an algorithm is initially given only one side of the vertex set of a bipartite graph; the vertices on the other side are revealed to the algorithm one by one, along with its incident edges. The algorithm is required to maintain a matching in the current graph, where the algorithm revises the matching after each vertex arrival by reassigning vertices. Bernstein, Holm, and Rotenberg showed that an online algorithm can maintain a matching of maximum cardinality by performing amortized $O(\log^2 n)$ reassignments per arrival. In this paper, we propose to consider the general question of how requiring a non-amortized hard budget $k$ on the number of reassignments affects the algorithms' performances, under various models from the literature. We show that a simple, widely-used algorithm is a best-possible deterministic algorithm for all these models. For the unweighted maximum-cardinality problem, the algorithm is a $(1-\frac{2}{k+2})$-competitive algorithm, which is the best possible for a deterministic algorithm both under vertex arrivals and edge arrivals. Applied to the load balancing problem, this yields a bifactor online algorithm. For the weighted problem, which is traditionally studied assuming the triangle inequality, we show that the power of reassignment allows us to lift this assumption and the algorithm becomes a $\frac{1}{2}$-competitive algorithm for $k=4$, improving upon the $\frac{1}{3}$ of the previous algorithm without reassignments. We show that this also is a best-possible deterministic algorithm.

cs.DS↗

arXiv2019

Constant-Factor Approximation Algorithms for Parity-Constrained Facility Location Problems

Facility location is a prominent optimization problem that has inspired a large quantity of both theoretical and practical studies in combinatorial optimization. Although the problem has been investigated under various settings reflecting typical structures within the optimization problems of practical interest, little is known on how the problem behaves in conjunction with parity constraints. This shortfall of understanding was rather disturbing when we consider the central role of parity in the field of combinatorics. In this paper, we present the first constant-factor approximation algorithm for the facility location problem with parity constraints. We are given as the input a metric on a set of facilities and clients, the opening cost of each facility, and the parity requirement--odd, even, or unconstrained--of every facility in this problem. The objective is to open a subset of facilities and assign every client to an open facility so as to minimize the sum of the total opening costs and the assignment distances, but subject to the condition that the number of clients assigned to each open facility must have the same parity as its requirement. Although the unconstrained facility location problem as a relaxation for this parity-constrained generalization has unbounded gap, we demonstrate that it yields a structured solution whose parity violation can be corrected at small cost. This correction is prescribed by a T-join on an auxiliary graph constructed by the algorithm. This graph does not satisfy the triangle inequality, but we show that a carefully chosen set of shortcutting operations leads to a cheap and sparse T-join. Finally, we bound the correction cost by exhibiting a combinatorial multi-step construction of an upper bound. We also present the first constant-factor approximation algorithm for the parity-constrained k-center problem, the bottleneck optimization variant.

cs.DS↗