Dongyang Wang

About Me

I am a research assistant and M.S. student in Computer Science at Monmouth University, with a strong foundation in both computer science and finance. I hold an M.S. in Financial Engineering from Rensselaer Polytechnic Institute (RPI) and dual B.S. degrees in Finance and Law from East China University of Science and Technology.

My research emphasizes empirical applications of machine learning to solve real-world problems. I focus on developing novel architectures, particularly Time-Aware Transformers for medical time-series prediction, and exploring the intersection of quantum computing and machine learning. I'm passionate about creating interpretable, efficient AI systems with practical impact in healthcare and financial domains.

Beyond research, I enjoy playing pickleball and painting landscapes, which help me maintain creativity and balance in my academic journey.

Research Projects

Time-Aware Transformer for AECOPD Prediction

Developed a novel Time-Aware Transformer that uses temporal embeddings to model long-range dependencies in ventilator data for predicting respiratory deterioration in COPD patients. Achieved R² of 0.78 on a 2-billion-row dataset.

Deep Learning Healthcare AI Transformers

Multimodal Injury Risk Prediction in Tennis

Created a weighted ensemble learning framework integrating wearable sensors, wellness scores, and video analysis to predict athlete injury risk and performance. Novel approach to handling multimodal sports data.

Ensemble Learning Multimodal AI Sports Analytics

Quantum Machine Learning

Investigating Quantum Transformers and QNN architectures, comparing their performance against GPU-based systems across speed, accuracy, and power consumption. Exploring applications in financial time-series prediction.

Quantum Computing QML Time Series

Publications

Time-Aware Transformer-based Prediction Model for AECOPD

Qu, W., Zheng, L., Wang, D., Wang, J., & Pan, H. (2025)

Studies in Health Technology and Informatics, 329, 1089-1093

View Publication →

Multimodal Injury Risk and Performance Prediction in Tennis using Weighted Ensemble Learning

Qu, W., Wang, D., Zheng, L., Alvarez, F. E., Polasa, S., & Wang, J.

Under Review at Communications of the ACM

Time-to-Event Forecasting of AECOPD using a Time-Aware Transformer

Wang, D., Qu, W., Zheng, L., & Wang, J.

In Preparation

Curriculum Vitae

Education

Monmouth University

West Long Branch, NJ

2024 - 2026 (Expected)

M.S. in Computer Science, GPA: 3.94

Rensselaer Polytechnic Institute (RPI)

Troy, NY

2012 - 2014

M.S. in Financial Engineering and Risk Analytics, Quantitative Finance Track, GPA: 3.75

East China University of Science and Technology

Shanghai, China

2008 - 2012

B.S. in Finance, GPA: 3.5

B.S. in Law, GPA: 3.5

Research Experience

Research Assistant

Monmouth University, West Long Branch, NJ

2024 - Present

Designed and implemented Time-Aware Transformer models, including both classification and regression-based time to event forecasting on large-scale ventilator datasets
Evaluated Quantum Machine Learning (QML) architectures and compared their performance to GPU-based models in terms of computational efficiency, predictive accuracy, and power usage
Developed multimodal weighted ensemble learning frameworks to model athlete injury risk and readiness score

Selected Course Projects

Deep Neural Networks for Hospital Length-of-Stay (LOS) Prediction

Developed and tuned a deep neural network using Keras Tuner to optimize architecture and hyperparameters

Sports Injury Sentinel LLM (PEFT + DPO Medical Chatbot)

Created a domain-specific sports-injury medical dataset and implemented LLM-as-Judge evaluations. Fine-tuned the Qwen-3 model using PEFT method (LoRA/QLoRA) and applied Direct Preference Optimization (DPO) to improve preference-aligned responses

LLM based Company Shareholder Letter Sentiment Analysis

Applied zero and few-shot learning with Qwen-3 to extract strategic themes and sentiment trends from shareholder letters

Precision Agriculture Image Classification (CNN + Transfer Learning)

Built multiple image-classification models (CNN and VGG16 transfer learning) to identify different plant species

Technical Skills

Programming: Python, Java, C++, MATLAB, R, SQL, MongoDB

Machine Learning: Deep Learning, Transformers, LLMs, Computer Vision, Time Series Analysis, Quantum ML

Research Areas: Healthcare AI, Financial AI, Ensemble Learning, Model Interpretability