Academic profile
These themes summarize completed M.Sc. work and the doctoral directions I will continue at Xiamen University, while keeping documented outputs distinct from forward-looking research plans.
Recent M.Sc. work centered on multimodal forecasting, non-stationary time-series analysis, and early-warning-oriented modelling for infectious diseases, especially respiratory disease activity and influenza-related scenarios.
A central line of work studies co-circulation patterns, synchronous and lagged relationships, and feature extraction for influenza-related signals in China, using interpretable time-series and multiscale analytical methods.
Recent work on the 2025 Foshan chikungunya outbreak combines ODE and Petri Net formulations to compare transmission interpretation, intervention effects, and uncertainty under small-sample conditions.
My broader research narrative links macroscopic epidemic patterns with microscopic viral evolution, and the longer-term agenda extends toward connecting epidemiological signals, phylodynamic evidence, and future early-warning models within a coherent framework.
The current phase connects completed M.Sc. modelling work, representative evidence, and the next doctoral stage in epidemiology and health statistics.
My completed M.Sc. and continuing doctoral trajectory includes MAESTRO for multimodal respiratory disease forecasting, influenza co-circulation and co-infection analysis, dual-framework ODE/Petri Net modelling for chikungunya transmission, and ongoing maintenance of an infectious-disease intelligence platform for data collection and analytical support. The broader narrative focuses on non-stationary time-series analysis, small-sample outbreak dynamics, and linking macroscopic epidemic patterns with microscopic viral evolution in future early-warning systems. I aim to continuously integrate epidemiological signals with phylodynamic evidence to provide reliable, multi-scale insights for public health preparedness and decision-making.
Developed the MAESTRO multimodal forecasting framework for respiratory disease activity; in the documented evaluation context, the reported R² reached 0.956 on a 10-year Hong Kong influenza dataset.
Built an ODE and Petri Net dual-model workflow for the 2025 Foshan chikungunya outbreak to compare intervention phases, transmission indicators, and sensitivity under small-sample conditions.
Designed an analytical pipeline for influenza co-circulation and co-infection signals, leveraging interpretable time-series decomposition and multiscale frequency-domain coupling patterns.
Independently developed and continuously maintained an infectious-disease news collection, database, and visualization platform that supports research-oriented data acquisition and monitoring workflows.
Methodologically, I try to connect data sources, model structure, interpretation, and reproducible implementation while distinguishing documented evidence from forward-looking doctoral agenda items.
I work with surveillance records, environmental information, behavioral or news-derived signals, and, in longer-term plans, genomic data, aiming to turn heterogeneous public-health information into coherent analytical inputs.
My current work combines decomposition-based forecasting, state-space reasoning, and mechanistic models such as ODEs and Petri Nets so that predictive results remain tied to epidemiological interpretation.
I distinguish completed outputs, work under review, work in preparation, software artifacts, and forward-looking doctoral plans, so that the research narrative stays aligned with documented evidence.