Overview
- Develops and fits Hawkes process models to healthcare event data.
- Uses density power divergence ideas to build more robust inference procedures.
- Establishes a relationship between Hawkes processes and integer-valued autoregression models.
Research Project
Private and Robust Inference for Hawkes Process Data and Integer-Valued Autoregression
Robust and privacy-aware inference for Hawkes process data, with a connection to integer-valued autoregression and applications to healthcare event streams.
In Process
Hawkes Processes
Healthcare Data
Methods and contribution
The project studies self-exciting temporal point processes in settings where anomalies or model misspecification can distort standard inference. Robustness is introduced through divergence-based estimation, and numerical experiments are carried out using thinning-based simulation.
By linking Hawkes process structure to integer-valued autoregression, the work also helps connect point-process modeling with familiar discrete-time count models.