Existence and non-existence of consistent estimators in supercritical controlled branching processes
By: Peter Braunsteins, Sophie Hautphenne, James Kerlidis
Potential Business Impact:
Figures out how populations grow and change.
We consider the problem of estimating the parameters of a supercritical controlled branching process consistently from a single observed trajectory of population size counts. Our goal is to establish which parameters can and cannot be consistently estimated. When a parameter can be consistently estimated, we derive an explicit expression for the estimator. We address these questions in three scenarios: when the distribution of the control function distribution is known, when it is unknown, and when progenitor numbers are observed alongside population size counts. Our results offer a theoretical justification for the common practice in population ecology of estimating demographic and environmental stochasticity using separate observation schemes.
Similar Papers
Consistent estimation in subcritical birth-and-death processes
Statistics Theory
Helps predict how populations grow or shrink.
Gaussian approximations for fast Bayesian inference of partially observed branching processes with applications to epidemiology
Methodology
Makes tracking disease spread much faster.
On Privately Estimating a Single Parameter
Machine Learning (CS)
Keeps personal data safe while still using it.