arXiv subjects

Qianwen Guo

Publications and source records attributed to Qianwen Guo.

8 recordsLinked to original sources

arXiv2026

An Embarrassingly Simple Detector for Model Extraction Attacks in Large Language Model API Traffic

Large language models (LLMs) are increasingly deployed through hosted APIs, making model extraction a practical threat to model ownership and service security. Individual extraction queries often resemble benign requests, while existing evaluations often focus on single-query anomaly scoring or pure benign-versus-attacker user settings. We formulate model extraction monitoring as benign-calibrated traffic-window distribution testing: embed incoming queries into a semantic space and test whether their aggregate distribution deviates from historical benign traffic. We instantiate this formulation with maximum mean discrepancy (MMD), using only benign-vs-benign comparisons to set the decision threshold. We evaluate on fourteen attacker-normal query pairs from four extraction scenarios and compare with adapted PRADA, SEAT, CAP, DATE, marginal Mahalanobis, and pseudo-class energy baselines. Across three random seeds, MMD achieves 0.3% benign FPR, 100.0% pure-attacker TPR, 90.5% average TPR over attacker fractions, and 95.1% balanced accuracy. These results show that benign-calibrated distribution testing is a strong empirical baseline for model extraction detection in both user-level and mixed multi-user LLM API traffic.

cs.CR↗

arXiv2025

Measuring Nonlinear Relationships and Spatial Heterogeneity of Influencing Factors on Traffic Crash Density Using GeoXAI

This study applies a Geospatial Explainable AI (GeoXAI) framework to analyze the spatially heterogeneous and nonlinear determinants of traffic crash density in Florida. By combining a high-performing machine learning model with GeoShapley, the framework provides interpretable, tract-level insights into how roadway characteristics and socioeconomic factors contribute to crash risk. Specifically, results show that variables such as road density, intersection density, neighborhood compactness, and educational attainment exhibit complex nonlinear relationships with crashes. Extremely dense urban areas, such as Miami, show sharply elevated crash risk due to intensified pedestrian activities and roadway complexity. The GeoShapley approach also captures strong spatial heterogeneity in the influence of these factors. Major metropolitan areas including Miami, Orlando, Tampa, and Jacksonville display significantly higher intrinsic crash contributions, while rural tracts generally have lower baseline risk. Each factor exhibits pronounced spatial variation across the state. Based on these findings, the study proposes targeted, geography-sensitive policy recommendations, including traffic calming in compact neighborhoods, adaptive intersection design, speed management on high-volume corridors such as I-95 in Miami, and equity-focused safety interventions in disadvantaged rural areas of central and northern Florida. Moreover, this paper compares the results obtained from GeoShapley framework against other established methods (e.g., SHAP and MGWR), demonstrating its powerful ability to explain nonlinearity and spatial heterogeneity simultaneously.

stat.ME↗

arXiv2025

Dynamic Switching Models for Truck-only Delivery and Drone-assisted Truck Delivery under Demand Uncertainty

Integrating drones into truck delivery systems can improve customer accessibility, reduce operational costs, and increase delivery efficiency. However, drone deployment incurs costs, including procurement, maintenance, and energy consumption, and its benefits depend on service demand. In low-demand areas, drone-assisted trucks may underutilize resources due to high upfront costs. Accurately predicting demand is challenging due to uncertainties from unforeseen events or infrastructure disruptions. To address this, a market entry and exit real option approach is used to optimize switching between truck-only and drone-assisted delivery under stochastic demand. Results show that deploying multiple drones per truck offers significant cost advantages in high-demand regions. Using the proposed dynamic switching model, deterministic and stochastic approaches reduce costs by 17.4% and 31.3%, respectively, compared to immediate cost-saving switching. Sensitivity analysis reveals asymmetric effects of stochastic parameters on entry and exit timings. A stochastic multiple-options model is further developed to dynamically switch between truck-only and drone-assisted delivery with varying drone numbers. Applying these models to Miami-Dade County, we evaluate dynamic switching costs for three major logistics operators. This study highlights the potential benefits of dynamic delivery switching and provides insights for optimizing logistics operations.

eess.SY↗

arXiv2025

VRISE: A Virtual Reality Platfrom for Immersive and Interactive Surveying Education

Surveying is a core component of civil engineering education, requiring students to engage in hands-on spatial measurement, instrumentation handling, and field-based decision-making. However, traditional instruction often poses logistical and cognitive challenges that can hinder accessibility and student engagement. While virtual laboratories have gained traction in engineering education, few are purposefully designed to support flexible, adaptive learning in surveying. To address this gap, we developed Virtual Reality for Immersive and Interactive Surveying Education (VRISE), an immersive virtual reality laboratory that replicates ground-based and aerial surveying tasks through customizable, accessible, and user-friendly modules. VRISE features interactive experiences such as differential leveling with a digital level equipment and waypoint-based drone navigation, enhanced by input smoothing, adaptive interfaces, and real-time feedback to accommodate diverse learning styles. Evaluation across multiple user sessions demonstrated consistent gains in measurement accuracy, task efficiency, and interaction quality, with a clear progression in skill development across the ground-based and aerial surveying modalities. By reducing cognitive load and physical demands, even in tasks requiring fine motor control and spatial reasoning, VRISE demonstrates the potential of immersive, repeatable digital environments to enhance surveying education, broaden participation, and strengthen core competencies in a safe and engaging setting.

cs.HC↗

arXiv2025

Optimized dynamic scheduling of an exclusive bus lane or high occupancy vehicle lane in a bimodal traffic corridor

Efficient management of traffic corridors is critical for sustaining urban mobility. Exclusive bus lane (EBL) and high occupancy vehicle lane (HOVL) are two prominent strategies for enhancing public transit services and alleviating congestion. EBLs prioritize bus transit by providing dedicated lanes for faster travel times, while HOVLs encourage carpooling by reserving lanes for high-occupancy vehicles. However, static implementations of these policies may underutilize road resources and disrupt general-purpose lanes. Dynamic implementation, based on real-time demand, can potentially maximize road efficiency and minimize negative impacts. This study first compares the mixed traffic policy (MTP), exclusive bus lane policy (EBLP), and high occupancy vehicle lane policy (HOVLP) by formulating the total system costs in the context of a bimodal traffic corridor involving private cars and public buses. Under each lane policy, the bus frequency is optimized together with the modal split equilibrium derived separately. Based on dynamic demand simulated using an Ornstein-Uhlenbeck (O-U) process, switching thresholds are then derived to identify optimal periods for implementing each policy. Results reveal significant reductions in total system costs with the proposed dynamic policy schedules. Compared to static implementations, the dynamic policy schedules achieve cost reductions of 11.9%, 6.70%, and 43.64% relative to MTP-only, EBLP-only, and HOVLP-only scenarios, respectively. Additionally, in two real case studies of existing EBL and HOVL operations in Seattle, the proposed dynamic policy reduces total costs by 32.5% and 28.6%, respectively. The findings provide valuable insights for policymakers and transit planners, offering a robust framework for dynamically scheduling and integrating EBL and HOVL policies to optimize urban corridor efficiency and reduce overall system costs.

math.OC↗

arXiv2025

Equity-aware Design and Timing of Fare-free Transit Zoning under Demand Uncertainty

We propose the first analytical stochastic model for optimizing the configuration and implementation policies of fare-free transit. The model focuses on a transportation corridor with two transportation modes: automobiles and buses. The corridor is divided into two sections, an inner one with fare-free transit service and an outer one with fare-based transit service. Under the static version of the model, the optimized length and frequency of the fare-free transit zone can be determined by maximizing total social welfare. The findings indicate that implementing fare-free transit can increase transit ridership and reduce automobile use within the fare-free zone while social equity among the demand groups can be enhanced by lengthening the fare-free zone. Notably, the optimal zone length increases when both social welfare and equity are considered jointly, compared to only prioritizing social welfare. The dynamic model, framed within a market entry and exit real options approach, solves the fare policy switching problem, establishing optimal timing policies for activating or terminating fare-free service. The results from dynamic models reveal earlier implementation and extended durations of fare-free transit in the social welfare-aware regime, driven by lower thresholds compared to the social equity-aware regime.

math.OC↗

arXiv2025

Runway capacity expansion planning for public airports under demand uncertainty

Flight delay is a significant issue affecting air travel. The runway system, frequently falling short of demand, serves as a bottleneck. As demand increases, runway capacity expansion becomes imperative to mitigate congestion. However, the decision to expand runway capacity is challenging due to inherent uncertainties in demand forecasts. This paper presents a novel approach to modeling air traffic demand growth as a jump diffusion process, incorporating two layers of uncertainty: Geometric Brownian Motion (GBM) for continuous variability and a Poisson process to capture the impact of crisis events, such as natural disasters or public health emergencies, on decision-making. We propose a real options model to jointly evaluate the interrelated factors of optimal runway capacity and investment timing under uncertainty, with investment timing linked to trigger demand. The findings suggest that increased uncertainty indicates more conservative decision-making. Furthermore, the relationship between optimal investment timing and expansion size is complex: if the expansion size remains unchanged, the trigger demand decreases as the demand growth rate increases; if the expansion size experiences a jump, the trigger demand also exhibits a sharp rise. This work provides valuable insights for airport authorities for informed capacity expansion decision-making.

eess.SY↗

arXiv2025

Vehicle occupancy estimation in Automated Guideway Transit via deep learning with Wi-Fi probe requests

This study contributes to the advancement of vehicle occupancy estimation in Automated Guideway Transit (AGT) systems using Wi-Fi probe requests and deep learning models. We propose a comprehensive framework for evaluating various approaches to occupancy estimation, particularly in the context of MAC address randomization. While many methods proposed in the literature claim effectiveness in simpler experimental settings, our research reveals that those methods are unreliable in the complex environment of AGT systems. Specifically, techniques for handling randomized MAC addresses and distinguishing between passenger and non-passenger data do not perform well in AGT systems. Despite challenges in tracking individual devices, our study demonstrates that accurate occupancy estimation using Wi-Fi probe requests remains feasible. A pilot study conducted on the Miami-Dade Metromover, an AGT system characterized by frequent stops, significant occupancy fluctuations, and absence of fare collection devices, provides a robust testing ground for the framework. Additionally, our findings show that deep learning models significantly outperform machine learning models in this context. The insights from this study can significantly enhance decision-making for transit agencies to optimize operations and elevate service quality.

eess.SP↗