Dimy Anguima Ibondzi, Herve2014-04-162014-04-1620142014http://hdl.handle.net/10393/30906http://dx.doi.org/10.20381/ruor-3637In this thesis we are concerned with estimation of clustering probabilities for univariate heavy tailed time series. We employ functional convergence of a bivariate tail empirical process to conclude asymptotic normality of an estimator of the clustering probabilities. Theoretical results are illustrated by simulation studies.enConditional ProbabilitiesThesis