arXiv subjects

Wentao Weng

Publications and source records attributed to Wentao Weng.

9 recordsLinked to original sources

arXiv2025

Learning to Defer in Congested Systems: The AI-Human Interplay

High-stakes applications rely on combining Artificial Intelligence (AI) and humans for responsive and reliable decision making. For example, content moderation in social media platforms often employs an AI-human pipeline to promptly remove policy violations without jeopardizing legitimate content. A typical heuristic estimates the risk of incoming content and uses fixed thresholds to decide whether to auto-delete the content (classification) and whether to send it for human review (admission). This approach can be inefficient as it disregards the uncertainty in AI's estimation, the time-varying element of content arrivals and human review capacity, and the selective sampling in the online dataset (humans only review content filtered by the AI). In this paper, we introduce a model to capture such an AI-human interplay. In this model, the AI observes contextual information for incoming jobs, makes classification and admission decisions, and schedules admitted jobs for human review. During these reviews, humans observe a job's true cost and may overturn an erroneous AI classification decision. These reviews also serve as new data to train the AI but are delayed due to congestion in the human review system. The objective is to minimize the costs of eventually misclassified jobs. We propose a near-optimal learning algorithm that carefully balances the classification loss from a selectively sampled dataset, the idiosyncratic loss of non-reviewed jobs, and the delay loss of having congestion in the human review system. To the best of our knowledge, this is the first result for online learning in contextual queueing systems. Moreover, numerical experiments based on online comment datasets show that our algorithm can substantially reduce the number of misclassifications compared to existing content moderation practice.

cs.LG↗

arXiv2025

Scheduling in Queueing Systems with Uncertain and Evolving Holding Costs

In content moderation for social media platforms, the cost of delaying the review of a content is proportional to its view trajectory, which fluctuates and is apriori unknown. Motivated by such uncertain and evolving holding costs, we consider a queueing model where job states evolve based on a Markov chain with state-dependent instantaneous holding costs. We demonstrate that in the presence of such uncertain and evolving holding costs, the two canonical algorithmic principles, instantaneous-cost ($c\mu$-rule) and expected-remaining-cost ($c\mu/\theta$-rule), are suboptimal. By viewing each job as a Markovian ski-rental problem, we develop a new index-based algorithm, Opportunity-adjusted Remaining Cost (OaRC), that adjusts to the opportunity of serving jobs in the future when uncertainty partly resolves. We show that the suboptimality gap of OaRC scales as $\tilde{O}(\sqrt{N})$, where $N$ is the system size. This bound shows that OaRC achieves asymptotic optimality for overloaded systems when the system size $N$ scales to infinity. Moreover, the bound is independent of the state-space size, which is a desirable property when job states contain contextual information. We corroborate our results with an extensive simulation study based on two holding cost patterns (online ads and user-generated content) that arise in content moderation for social media platforms. Our simulations based on synthetic and real datasets demonstrate that OaRC consistently outperforms existing practice, which is based on the two canonical algorithmic principles.

cs.DS↗

arXiv2025

The Transient Cost of Learning in Queueing Systems

Queueing systems are widely applicable stochastic models with use cases in communication networks, healthcare, service systems, etc. Although their optimal control has been extensively studied, most existing approaches assume perfect knowledge of the system parameters. This assumption rarely holds in practice where there is parameter uncertainty, thus motivating a recent line of work on bandit learning for queueing systems. This nascent stream of research focuses on the asymptotic performance of the proposed algorithms but does not provide insight on the transient performance in the early stages of the learning process. In this paper, we propose the Transient Cost of Learning in Queueing (TCLQ), a new metric that quantifies the maximum increase in time-averaged queue length caused by parameter uncertainty. We characterize the TCLQ of a single-queue multi-server system, and then extend these results to multi-queue multi-server systems and networks of queues. In establishing our results, we propose a unified analysis framework for TCLQ that bridges Lyapunov and bandit analysis, provides guarantees for a wide range of algorithms, and could be of independent interest.

cs.LG↗

arXiv2025

Group fairness in dynamic refugee assignment

Ensuring that refugees and asylum seekers thrive (e.g., find employment) in their host countries is a profound humanitarian goal, and a primary driver of employment is the geographic location within a host country to which the refugee or asylum seeker is assigned. Recent research has proposed and implemented algorithms that assign refugees and asylum seekers to geographic locations in a manner that maximizes the average employment across all arriving refugees. While these algorithms can have substantial overall positive impact, using data from two industry collaborators we show that the impact of these algorithms can vary widely across key subgroups based on country of origin, age, or educational background. Thus motivated, we develop a simple and interpretable framework for incorporating group fairness into the dynamic refugee assignment problem. In particular, the framework can flexibly incorporate many existing and future definitions of group fairness from the literature (e.g., Random, Proportionally Optimized within-group, and MaxMin). Equipped with our framework, we propose two bid-price algorithms that maximize overall employment while simultaneously yielding provable group fairness guarantees. Through extensive numerical experiments using various definitions of group fairness and real-world data from the U.S. and the Netherlands, we show that our algorithms can yield substantial improvements in group fairness compared to an offline benchmark fairness constraints, with only small relative decreases ($\approx$ 1%-5%) in global performance.

cs.GT↗

arXiv2023

Efficient decentralized multi-agent learning in asymmetric bipartite queueing systems

We study decentralized multi-agent learning in bipartite queueing systems, a standard model for service systems. In particular, N agents request service from K servers in a fully decentralized way, i.e, by running the same algorithm without communication. Previous decentralized algorithms are restricted to symmetric systems, have performance that is degrading exponentially in the number of servers, require communication through shared randomness and unique agent identities, and are computationally demanding. In contrast, we provide a simple learning algorithm that, when run decentrally by each agent, leads the queueing system to have efficient performance in general asymmetric bipartite queueing systems while also having additional robustness properties. Along the way, we provide the first provably efficient UCB-based algorithm for the centralized case of the problem.

cs.LG↗

arXiv2022

The Mean-Squared Error of Double Q-Learning

In this paper, we establish a theoretical comparison between the asymptotic mean-squared error of Double Q-learning and Q-learning. Our result builds upon an analysis for linear stochastic approximation based on Lyapunov equations and applies to both tabular setting and with linear function approximation, provided that the optimal policy is unique and the algorithms converge. We show that the asymptotic mean-squared error of Double Q-learning is exactly equal to that of Q-learning if Double Q-learning uses twice the learning rate of Q-learning and outputs the average of its two estimators. We also present some practical implications of this theoretical observation using simulations.

cs.LG↗

arXiv2021

Labor-right Protecting Dispatch of Meal Delivery Platforms

The boom in the meal delivery industry brings growing concern about the labor rights of riders. Current dispatch policies of meal-delivery platforms focus mainly on satisfying consumers or minimizing the number of riders for cost savings. There are few discussions on improving the working conditions of riders by algorithm design. The lack of concerns on labor rights in mechanism and dispatch design has resulted in a very large time waste for riders and their risky driving. In this research, we propose a queuing-model-based framework to discuss optimal dispatch policy with the goal of labor rights protection. We apply our framework to develop an algorithm minimizing the waiting time of food delivery riders with guaranteed user experience. Our framework also allows us to manifest the value of restaurants' data about their offline-order numbers on improving the benefits of riders.

cs.PF↗

arXiv2020

Achieving Zero Asymptotic Queueing Delay for Parallel Jobs

Zero queueing delay is highly desirable in large-scale computing systems. Existing work has shown that it can be asymptotically achieved by using the celebrated Power-of-$d$-choices (pod) policy with a probe overhead $d = ω\left(\frac{\log N}{1-λ}\right)$, and it is impossible when $d = O\left(\frac{1}{1-λ}\right)$, where $N$ is the number of servers and $λ$ is the load of the system. However, these results are based on the model where each job is an indivisible unit, which does not capture the parallel structure of jobs in today's predominant parallel computing paradigm. This paper thus considers a model where each job consists of a batch of parallel tasks. Under this model, we propose a new notion of zero (asymptotic) queueing delay that requires the job delay under a policy to approach the job delay given by the max of its tasks' service times, i.e., the job delay assuming its tasks entered service right upon arrival. This notion quantifies the effect of queueing on a job level for jobs consisting of multiple tasks, and thus deviates from the conventional zero queueing delay for single-task jobs in the literature. We show that zero queueing delay for parallel jobs can be achieved using the batch-filling policy (a variant of the celebrated pod policy) with a probe overhead $d = ω\left(\frac{1}{(1-λ)\log k}\right)$ in the sub-Halfin-Whitt heavy-traffic regime, where $k$ is the number of tasks in each job { and $k$ properly scales with $N$ (the number of servers)}. This result demonstrates that for parallel jobs, zero queueing delay can be achieved with a smaller probe overhead. We also establish an impossibility result: we show that zero queueing delay cannot be achieved if $d = \exp\left({o\left(\frac{\log N}{\log k}\right)}\right)$.

cs.PF↗

arXiv2020

Optimal Load Balancing in Bipartite Graphs

Applications in cloud platforms motivate the study of efficient load balancing under job-server constraints and server heterogeneity. In this paper, we study load balancing on a bipartite graph where left nodes correspond to job types and right nodes correspond to servers, with each edge indicating that a job type can be served by a server. Thus edges represent locality constraints, i.e., each job can only be served at servers which contained certain data and/or machine learning (ML) models. Servers in this system can have heterogeneous service rates. In this setting, we investigate the performance of two policies named Join-the-Fastest-of-the-Shortest-Queue (JFSQ) and Join-the-Fastest-of-the-Idle-Queue (JFIQ), which are simple variants of Join-the-Shortest-Queue and Join-the-Idle-Queue, where ties are broken in favor of the fastest servers. Under a "well-connected" graph condition, we show that JFSQ and JFIQ are asymptotically optimal in the mean response time when the number of servers goes to infinity. In addition to asymptotic optimality, we also obtain upper bounds on the mean response time for finite-size systems. We further show that the well-connectedness condition can be satisfied by a random bipartite graph construction with relatively sparse connectivity.

cs.PF↗