arXiv subjects

Daniel Zhuoyu Long

Publications and source records attributed to Daniel Zhuoyu Long.

7 recordsLinked to original sources

arXiv2026

Selective Ambulance Dispatch Under Contextual Travel-Time Uncertainty

Ambulance response is time-critical in out-of-hospital cardiac arrest (OHCA), where dispatchers must balance timely arrivals with limited fleet capacity. Static territories and deterministic travel-time estimates are vulnerable to dynamic congestion, while always-dual dispatch adds redundancy but consumes fleet capacity. We propose IDEAL (Intelligent Dual dispatch of Emergency AmbuLances), a selective dual-dispatch framework that sends a second ambulance only when the optimistic gap between primary and secondary paths exceeds a threshold. IDEAL learns context-specific edge travel times from trip-level dispatch records, including unobserved routes, using a weakly supervised bilevel representation network. We train the nonsmooth model with mini-batch conservative gradients and prove an asymptotic convergence guarantee. IDEAL models uncertainty via Burg-divergence perturbations to a shared metric in the learned representation space, thereby inducing correlated changes in edge travel times and learning context-specific radii from historical underprediction errors. For real-time decisions, IDEAL casts optimistic-gap computation as a difference-of-convex program and derives an efficient oracle with complexity guarantees. In collaboration with the Hong Kong Fire Services Department, we evaluate IDEAL using historical OHCA records and real-time adaptive simulations. The results achieve a stronger response-time/resource trade-off relative to all region-based and Google-based baselines.

math.OC↗

arXiv2026

From Optimization to Satisficing: Robust Screening under Distributional Ambiguity

This study investigates a robust screening problem under distributional ambiguity, where a seller is uncertain about a buyer's true valuation distribution, knowing only that it lies near a reference distribution measured by the Wasserstein metric. Traditional robust optimization (RO) approaches prioritize maximizing worst-case revenue within predefined ambiguity sets, often yielding seller-centric outcomes and reliance on precise set specifications. We propose a robust satisficing (RS) framework aimed at attaining a specified revenue target by minimizing the worst-case shortfall across all potential distributions. Our approach offers a tractable formulation and detailed characterization of optimal mechanisms using randomized pricing strategies. We also assess the out-of-sample efficacy of a simple posted pricing mechanism, finding it particularly effective with lower targets and positively skewed valuations, where smaller valuations have high probability mass. Comparing RO with RS, we find that RS consistently enhances buyer surplus when the reference distribution has an increasing hazard rate and increases out-of-sample seller revenue with positively skewed true valuations. Our analysis indicates that a target-driven RS framework enhances buyer surplus and fairness by offering more opportunities to lower-valuation buyers, potentially boosting overall revenue in scenarios with demand skewed toward these valuations. This approach offers a practical and viable modeling alternative to conventional RO methods, effectively overcoming the challenges of ambiguity set calibration while ensuring broader equitable access for diverse buyers.

math.OC↗

arXiv2026

Fragility-aware Classification for Understanding Risk and Improving Generalization

Classification models play a central role in data-driven decision-making applications such as medical diagnosis, recommendation systems, and risk assessment. Traditional performance metrics, such as accuracy and AUC, focus on overall error rates but fail to account for the confidence of incorrect predictions, i.e., the risk of confident misjudgments. This limitation is particularly consequential in safety-critical and cost-sensitive settings, where overconfident errors can lead to severe outcomes. To address this issue, we propose the Fragility Index (FI), a novel performance metric that evaluates classifiers from a risk-averse perspective by capturing the tail risk of confident misjudgments. We formulate FI within a robust satisficing (RS) framework to ensure robustness under distributional uncertainty. Building on this, we develop a tractable training framework that directly targets FI via a surrogate loss, and show that models trained under this framework admit provable bounds on FI. We further derive exact reformulations for a broad class of loss functions, including cross-entropy, hinge-type, and Lipschitz losses, and extend the approach to deep neural networks. Empirical results on real-world medical diagnosis tasks demonstrate that FI complements existing metrics by revealing error tail risk and improving decision quality. FI-based models achieve competitive accuracy and AUC while consistently reducing confident misjudgments and associated operational costs, offering a practical tool for improving robustness and reliability in risk-critical applications.

cs.LG↗

arXiv2025

Harmonizing SAA and DRO

Decision-makers often encounter uncertainty, and the distribution of uncertain parameters plays a crucial role in making reliable decisions. However, complete information is rarely available. The sample average approximation (SAA) approach utilizes historical data to address this, but struggles with insufficient data. Conversely, moment-based distributionally robust optimization (DRO) effectively employs partial distributional information but can yield conservative solutions even with ample data. To bridge these approaches, we propose a novel method called harmonizing optimization (HO), which integrates SAA and DRO by adaptively adjusting the weights of data and information based on sample size N. This allows HO to amplify data effects in large samples while emphasizing information in smaller ones. More importantly, HO performs well across varying data sizes without needing to classify them as large or small. We provide practical methods for determining these weights and demonstrate that HO offers finite-sample performance guarantees, proving asymptotic optimality when the weight of information follows a 1/\sqrt{N}-rate. In addition, HO can be applied to enhance scenario reduction, improving approximation quality and reducing completion time by retaining critical information from reduced scenarios. Numerical results show significant advantages of HO in solution quality compared to Wasserstein-based DRO, and highlight its effectiveness in scenario reduction.

math.OC↗

arXiv2024

Asymptotically Optimal Distributionally Robust Solutions through Forecasting and Operations Decentralization

Two-stage risk-averse distributionally robust optimization (DRO) problems are ubiquitous across many engineering and business applications. Despite their promising resilience, two-stage DRO problems are generally computationally intractable. To address this challenge, we propose a simple framework by decentralizing the decision-making process into two specialized teams: forecasting and operations. This decentralization aligns with prevalent organizational practices, in which the operations team uses the information communicated from the forecasting team as input to make decisions. We formalize this decentralized procedure as a bilevel problem to design a communicated distribution that can yield asymptotic optimal solutions to original two-stage risk-averse DRO problems. We identify an optimal solution that is surprisingly simple: The forecasting team only needs to communicate a two-point distribution to the operations team. Consequently, the operations team can solve a highly tractable and scalable optimization problem to identify asymptotic optimal solutions. Specifically, as the magnitude of the problem parameters (including the uncertain parameters and the first-stage capacity) increases to infinity at an appropriate rate, the cost ratio between our induced solution and the original optimal solution converges to one, indicating that our decentralized approach yields high-quality solutions. We compare our decentralized approach against the truncated linear decision rule approximation and demonstrate that our approach has broader applicability and superior computational efficiency while maintaining competitive performance. Using real-world sales data, we have demonstrated the practical effectiveness of our strategy. The finely tuned solution significantly outperforms traditional sample-average approximation methods in out-of-sample performance.

math.OC↗

arXiv2020

Scenario-decomposition Solution Framework for Nonseparable Stochastic Control Problems

When stochastic control problems do not possess separability and/or monotonicity, the dynamic programming pioneered by Bellman in 1950s fails to work as a time-decomposition solution method. Such cases have posted a great challenge to the control society in both theoretical foundation and solution methodologies for many years. With the help of the progressive hedging algorithm proposed by Rockafellar and Wets in 1991, we develop a novel scenario-decomposition solution framework for stochastic control problems which could be nonseparable and/or non-monotonic, thus extending the reach of stochastic optimal control. We discuss then some of its promising applications, including online quadratic programming problems and dynamic portfolio selection problems with smoothing properties.

math.OC↗

arXiv2019

Revised Progressive-Hedging-Algorithm Based Two-layer Solution Scheme for Bayesian Reinforcement Learning

Stochastic control with both inherent random system noise and lack of knowledge on system parameters constitutes the core and fundamental topic in reinforcement learning (RL), especially under non-episodic situations where online learning is much more demanding. This challenge has been notably addressed in Bayesian RL recently where some approximation techniques have been developed to find suboptimal policies. While existing approaches mainly focus on approximating the value function, or on involving Thompson sampling, we propose a novel two-layer solution scheme in this paper to approximate the optimal policy directly, by combining the time-decomposition based dynamic programming (DP) at the lower layer and the scenario-decomposition based revised progressive hedging algorithm (PHA) at the upper layer, for a type of Bayesian RL problem. The key feature of our approach is to separate reducible system uncertainty from irreducible one at two different layers, thus decomposing and conquering. We demonstrate our solution framework more especially via the linear-quadratic-Gaussian problem with unknown gain, which, although seemingly simple, has been a notorious subject over more than half century in dual control.

eess.SY↗