About
Zichun Zhu is currently a PhD candidate at the School of Electrical Engineering and Computer Science (EECS) at The University of Queensland (UQ). He is working under the supervision of A/Prof. Wen Hua, and A/Prof. Jiwon Kim. His current research interests include spatiotemporal data management, sequential modelling, map-matching, and AI/ML application on transportation.
Education
- Bachelor of Science in Electrical Systems - University of Melbourne (2015 - 2018)
- Master of Computer Science - University of Melbourne (2018 - 2020)
- Doctor of Philosophy - The University of Queensland (2021 - Present)
Publications
- Map-matching on Wireless Traffic Sensor Data with a Sequence-to-Sequence Model, MDM 2023
- CLMM: Uncertainty-aware Map-Matching for Bluetooth Data through Contrastive Learning, ADC 2024
- A Framework for Few-Shot Map-Matching through Semi-Supervised Self-Training, In progress
Projects
Infectious Disease (Malaria) Modelling (Mar 2019 - July 2020)
- Developed a population-level simulation model incorporating within-host parasite dynamics to investigate malaria transmission; Explored how within-host interactions influence broader epidemic patterns and transmission outcomes.
- Integrated stochastic simulation techniques and parameter calibration to ensure biological realism and model robustness; Contributed to understanding the lifecycle of malaria parasites and their implications for disease control strategies.
Bluetooth Map Matching (April 2021 - Present)
- ARC Linkage Project — Collaboration with Transport and Main Roads (TMR) & Brisbane City Council (BCC)
- Conducted three research studies on reconstructing individual trajectories from sparse, time-stamped Bluetooth records; Contributed to the ARC Linkage initiative to enhance real-time mobility intelligence and collaborated closely with TMR and BCC to align research with practical urban transport needs.
- Led data preprocessing, modeling, and evaluation to infer user-level trajectories; Developed a deep learning model with reinforcement learning to map trajectories to road network segments; Proposed contrastive learning framework to address data quality issues and improve trajectory consistency.
- The work has resulted in the publication of a paper, see Section Publications.
Contact
Email: zichun.zhu@uq.edu.au
LinkedIn: LinkedIn Profile
ORCID: 0000-0001-7986-4263
Google Scholar: Zichun Zhu