Agenda

Veronica Distefano

16 April 2020, 10 a.m

Abstract

In many scientific fields of research, recent advances in technology have allowed to gather spatial
data sets characterized by a very high number of variables. In this talk, I will present a new
clustering procedure for high-dimensional data in order to understand social and spatial divisions
in a geographical context. The novel procedure, called Bootstrap ClustGeo (BCG), is a hierarchical
approach based on bootstrap technique. More specifically, the BCG is derived by generating
multiple bootstrap clustering partitions and combining them with the Hamming distance. In this
seminar I will describe the new approach by presenting some case studies on simulated and real
data, showing the accuracy of its performance with the Internal Cluster Validity Measures (the
Silhouette Width, the Dunn’s and the Connectivity measures).

Bio sketch

Veronica Distefano has been a post-doctoral researcher at European Centre for Living Technology
contributing to develop research in spatial statistics mainly concerning the field of the MEDIUM EU
project. She graduated from the University of Salento and obtain the PhD in Statistics at the same
university. She was awarded a fellowship for a PostDoc position in that university for developing
research in computational statistical models for high dimensional data.