Introduction aux modèles espace-état et au filtre de Kalman

This article details the main concepts, problems and applications related to state-space models. We first present the general framework for these models and we, then, make explicit the algorithms used for the estimation, i.e. the Kalman filter and the EM algorithm. We, finally, analyse four applications : the decomposition of series into trend and cycle, the extraction of coincident economic indicators, the estimation of a time-varying equilibrium unemployment rate (TV-NAIRU) and the assessment of the informative content of the yield curve regarding future inflation.