Cox's Regression Model for Dynamics of Grouped Unemployment Data

The present contribution deals with models and statistical analysis of sequences of random events. The main characteristics of such a random point process is its intensity. We consider the case when observed data are grouped, i.e. when the numbers of observed events are summarized in discrete time periods. Simultaneously, the values of other factors influencing the intensity, the covariates, are grouped to distinct classes, too. The intensity of occurrence is modeled with the help of Cox's regression model, its formulation and the procedure of estimation of the parameters are presented. The approach is then applied to the analysis of unemployment data from period 1993 - 1999.