Events Archive

Abstract:

The Fleming-Viot process is a measure-valued diffusion that evolves under resampling (genetic drift), mutation, and selection. We consider interacting Fleming-Viot models for which the selection intensities are stochastic processes. We identify these measure-valued population models in fluctuating environments as the unique solutions of ``quenched'' martingale problems and present some convergence, ergodic, and averaging theorems for them. To this end, we extend the method of duality for time-inhomogeneous and quenched martingale problems.