Data Science

Session I

Arden Michel, "Broadband Internet Access as a Social Determinant During the Early COVID-19 Pandemic in Hampton Roads Cities and Counties"

This study aims to replicate and localize the 2025 study Broadband internet access as a social determinant in the early COVID-19 pandemic in US counties by Spencer Allen to the Hampton Roads area for reproducibility as well as apply the study with more robust broadband internet access data and standards. Data was drawn from publicly available datasets and analyzed using OLS regression. Results show a negative correlation between COVID-19 outcomes and mask usage with broadband internet access. Thus, the index study rings true under scrutiny, and as we move towards a more digitized world, broadband internet access can be viewed as a determinant for public health. 

Student Major(s): Computer Science 

Advisor: Dr. Trenton Ford

Zahra Rizvi, "Mapping Water Quality for Healthier Communities: the WaterWatch Mobile Application"

Urbanization impacts on water quality within the Chesapeake Bay's Hampton Roads area were quantified by correlating Total Suspended Matter (TSM) levels with population density (per 1,000 people) and urban land fraction (0-1). Landsat-8 and Sentinel-2 data were processed in the Open Data Cube framework on Google Colab, with ground truth measurements obtained from the Chesapeake Bay Data Hub. A multiple linear regression model (R^2 = 0.8) identified significant predictors: population (per 1,000) coefficient 0.04 mg/L, urban fraction coefficient 3.84 mg/L, and their interaction term (0.0028). Validation against in situ measurements yielded RMSE = 3.2 mg/L and bias = +1.1 mg/L. The WaterWatch mobile web application delivers location-based, monthly-updated TSM values to support environmental management and policy.

Student Major(s): Human Health & Physiology Major; Chemistry Minor

Advisor: Dr. Jennifer Swenson

Daniel Shi, "Re-SpS: Reinforcement Learning for Speculative Sampling"

Inference time latency has remained an open challenge for real world applications of large language models (LLMs). State-of-the-art (SOTA) speculative sampling (SpS) methods for LLMs, like EAGLE-3, use tree-based drafting to explore multiple candidate continuations in parallel. However, the hyperparameters controlling the tree structure are static, which limits flexibility and efficiency across diverse contexts and domains. We introduce Reinforcement learning for Speculative Sampling (Re-SpS), the first reinforcement learning (RL)-based framework for draft tree hyperparameter optimization. Re-SpS dynamically adjusts draft tree hyperparameters in real-time, learning context-aware policies that maximize generation speed by balancing speculative aggression with computational overhead. It leverages efficient state representations from target model hidden states and introduces multi-step action persistence for better context modeling. Evaluation results across five diverse benchmarks demonstrate consistent improvements over the SOTA method EAGLE-3, achieving up to 5.45× speedup over the backbone LLM and up to 1.12× speedup compared to EAGLE-3 across five diverse benchmarks, with no loss in output fidelity. Our code is included in the supplementary material and will be released upon paper acceptance.

Student Major(s): Data Science

Advisor: Dr. Haipeng Chen

Peiyan Xiao "Deployment of Dynamic Wireless Chargers to Enhance Electrification at the Port of Virginia"

The electrification of heavy-duty vehicles (HDVs) presents an opportunity to achieve significant economic and environmental benefits. However, this transition from HDVs to electric vehicles faces key challenges, including limited battery capacity, high ownership costs, and a lack of adequate charging infrastructure. Addressing these barriers is essential to making the electrification of HDVs viable and sustainable. One promising solution is the deployment of dynamic wireless chargers, which enable vehicles to charge while in motion, thereby extending their range and minimizing downtime. This study focuses on developing a comprehensive methodology for the strategic deployment of these chargers, considering factors such as practical feasibility, local demand, transportation logistics, and the compatibility of existing grid infrastructure. To validate this approach, the Port of Virginia serves as a case study, providing a real-world setting to demonstrate the feasibility and impact of HDV electrification. By addressing critical infrastructure gaps, this research aims to inform future electrification efforts, reduce greenhouse gas emissions, and enhance operational efficiency in transportation.

Student Major(s)/Minor: Data Science Major, Economics Minor

Advisor: Dr. Yanhai Xiong