Symmetry and Parameter Identifiability in Birth–Death–Sampling Models
- Discipline
- Applied mathematics; mathematical biology; infectious disease modelling
- Methodology
- Differential equations; structural identifiability analysis; group theory; symbolic computation; mathematical proof; simulation
- Date
- 30 Jul 2026
- Description
- Birth–death–sampling models can produce identical or nearly identical likelihoods under different combinations of transmission, removal and sampling parameters, making some epidemiological quantities fundamentally difficult to estimate. This project will use structural identifiability analysis and group-theoretic methods to characterise the transformations that leave model solutions or likelihoods unchanged. The student will identify parameter equivalence classes, determine which combinations of parameters are estimable from sampled transmission trees, and investigate how additional information or model constraints can break these symmetries. The results will provide a mathematical foundation for more reliable interpretation and parameterisation of phylodynamic models.
- Keywords
- Birth–death models;structural identifiability;symmetry;group theory
- Would Suit Applicants Who
- Have strong interests in applied mathematics, dynamical systems, algebra or mathematical epidemiology, and are comfortable with analytical and computational work
