A Tutorial on Observation-Driven Count Time Series Models Applied to Ecology
摘要
We study simple properties of observation-driven models for the analysis of count time series. Recent literature developed stochastic properties of such models together with associated inference. In addition the availability of software makes their fitting easily accessible. The tutorial offers an introduction to this research area by discussing well established models, their properties and associated inference for both univariate and multivariate count time series. Several other topics are also discussed and real data examples illustrate the methodology for analysis of ecological time series.