(submitted to Journal of Applied Science and Engineering, Jan–Oct 2025)

Under the supervision of Dr. Vo Thanh Ha, Faculty of Electrical and Electronics Engineering, University of Transport and Communications.

In this project, I helped build a deep learning model that combines RNN for time-series data and CNN for spatial data to predict daily asthma cases. We trained it with techniques like dropout, weight decay, early stopping, and data augmentation to make it more accurate and stable. Using real data from 2020–2024 in Seoul, Los Angeles, and Hanoi (including air quality, weather, and time factors), we evaluated our model with 5-fold cross-validation on a 24 GB GPU and shared the Python code publicly. Our CNN+GRU+Attention model achieved an AUC of 0.89 and an F1 score of 0.84, showing that combining AQI and weather data really improved the predictions.

(presented to The 10th International Invention Innovation Competition in Canada, iCAN 2025 1/2025 – 8/2025)

Using real data from 2020–2024 in Seoul, Los Angeles, and Hanoi, I helped process air quality, weather, and time-related factors to train our model. We built a hybrid deep learning approach where RNN handled time-series data and CNN handled spatial data, and we applied dropout, weight decay, early stopping, and data augmentation to make it more stable and accurate. After testing on a 24GB GPU with 5-fold cross-validation and releasing our Python code publicly, our CNN+GRU+Attention model achieved an AUC of 0.89 and an F1 score of 0.84 — showing how well AQI and weather data can work together for predicting asthma cases.

Submitted to Springer Journal | November 2024 – July 2025
Instructor: Assoc. Prof. Dr. Bui Phu Hung, Lecturer at Ton Duc Thang University

This chapter provides an overview of how AI can be applied to advance public health education, including creating personalized learning pathways, addressing human resource shortages through the use of extensive health-related data and digitalized health systems, and enabling simulation-based training that bridges the gap between theory and practice. It also discusses the automation of educational routines.

The research analyzes key challenges in applying AI to improve public health, such as equity and bias, explainability and trust, privacy and data security, infrastructure and workforce gaps, as well as legal and ethical constraints. It emphasizes the crucial role of institutions and organizations in fostering AI development and highlights the importance of establishing legal frameworks and policies to ensure the ethical and effective application of AI in healthcare practices.