[rede.APPIA] DaSSWeb – Data Science and Statistics Webinar – 28 May – Francesca Condino – Clustering Income Data Based on Share Densities

Zoom link

https://videoconf-colibri.zoom.us/j/94436835145

Starting from a situation where a reference population of income earners is naturally divided into subgroups, the aim is to explore the similarity of these sub-populations in terms of income inequality. To this end, a particular function, the so-called share density, strongly related to Lorenz curve and inequality measures, is considered for drawing information on income concentration. The Jensen-Shannon dissimilarity measure is proposed to evaluate the discrepancy across share densities, and some hierarchical and non-hierarchical clustering algorithms for unconventional data are defined. Obtained results on data from the Survey on Households Income and Wealth, carried out by Bank of Italy, and from EU-SILC survey, by Eurostat, are shown.