Classification of multiple short time series

The classification of multiple time series (MTS) in groups is an important issue that intersects many research fields. Although much research has been performed in modelling this type data, one area that has been largely overlooked is the particular type of time series where the data set consists of a large number of variables but with a small number of observations. The specific issue addressed in the present project concerns the possibility of classifying a set of administrative units (regions) in clusters on the basis of a new metric defined combining partial metrics computed on selected social and economic indicators that are only available for a limited period of time.