VIANEY LEOS BARAJAS

Abstract: Hidden Markov models (HMMs) are an increasingly popular statistical framework for the modeling of ecological time series as they assume that observations are manifestations of an underlying state process that functions as a proxy for behaviors of interest. For this reason, they have been used for identification of (unobserved) animal behaviors from telemetry data. However, fitting HMMs to data results in a mix of computational and statistical challenges not typically seen with other models. In this talk I will cover some of the most common challenges associated with fitting HMMs to ecological time series, how they can affect our inferences of ecological dynamics, but demonstrate why they are still a highly useful modeling framework.