Common Trends in European School Populations
Paola Sebastiani · 2001
This paper uses a novel Bayesian clustering method to categorize the temporal evolution of the share of population participating in tertiary/higher education in 14 European nations. The method represents time series as autoregressive models and applies an agglomerative clustering procedure to discover the most probable set of clusters describing the essential dynamics of these time series. To increase efficiency, the method uses a distance-based heuristic search strategy. This clustering method partitions the evolution of school population into three groups, thus revealing significant differences among tertiary/higher education in the 14 European nations.