ISAS provides undergraduate and graduate students with opportunities to participate in interdisciplinary research involving artificial intelligence, data assimilation, social computing, computational modeling, and physical AI.
Student projects:
Project Title: Rethinking Time-Series Forecasting as Binary Event Prediction with Spiking Neural Networks
Student: Md Tariqul Islam, a graduate student in Data Analytics, implemented and evaluated spiking neural network approaches using traffic time-series datasets. The project explored reframing time-series forecasting as binary event prediction.
Faculty Mentor: Peng Kang
The project explores an alternative approach to time-series analysis by focusing on whether an event will occur. The research was published at the 2026 IEEE International Conference on Networking and Communications (ICNC), with co-authors Peng Kang, Hivanshu Tyagi, and Md Tariqul Islam.
Project Title: Hybrid Scientific Machine Learning for Smart Building Thermal Modeling
Student: Graduate students Mit Pandya and Prince Chandra Singh developed a hybrid framework that combines physics-based thermal modeling, latent cloud-cover estimation, and Unscented Kalman Filtering to reconstruct unknown solar disturbances in smart buildings.
Faculty Mentor: Liang Kong
The project demonstrates the potential of Scientific Machine Learning to improve thermal-state estimation and support more energy-efficient building control. The research was accepted and to be published through IEEE CPESE 2026.
Ways to Participate
- Faculty-mentored research projects
- Independent study and capstone projects
- Graduate and undergraduate research opportunity
- Workshops, seminars, and research presentations
Express Your Interest
Students interested in participating should provide their academic program, research interests, relevant skills, and expected availability. Contact the faculty who you are interested in working with.