Please use this identifier to cite or link to this item: http://elea.unisa.it/xmlui/handle/10556/1439
Title: Il posizionamento dell’Acquedotto lucano spa: Una cluster analysis
Authors: Ciani, Angelo
Destefanis, Sergio Pietro
Destefanis, Sergio Pietro
Keywords: Cluster;Benchmark
Issue Date: 1-Apr-2014
Publisher: Universita degli studi di Salerno
Abstract: Research aims to analyze the operating performance of companies in the water services sector in the various Italian regions, to locate the position, in terms of overall profitability, of the Acquedotto Lucano Spa, a public company owned by the municipalities of the Region of Basilicata, which manages the integrated water service. Research was based on a sample where the database was constructed from data collected with the help of CNEL and the Osservatorio dei Servizi Pubblici Locali. It was, then, identified a set of indicators of financial statements that give an account of the economic structure and financial position of the various companies that manage the SII on the national territory. Through the use of appropriate techniques of clustering, we determined the distances between the statistical units in our sample and identified, finally, the positioning of the Region of Basilicata compared to other companies in the sample, obtaining, therefore, homogeneous groupings of companies that exhibit similar levels of economic, financial and equity indicators. Since we did not have data on the effectiveness of a particular clustering methodology compared to the other, we have followed different approaches to cluster analysis, thereby being able to measure how much each category of algorithms was appropriate to the study of such samples of data. Hierarchical methods have provided only general indications, a signal that the sample data are extremely cohesive and compact. a feature that has made problematic even the application of density methods, as it also determines a little variation in density between the data. Using k-means algorithm, we obtained important information about the number of clusters in which to split the sample, however, the values of the performance indicators we were not convinced of the complete reliability of the cluster boundaries identified. In order to obtain a more precise profile of the individual regions of homogeneity we have used a genetic algorithm, suitably calibrated, which has provided the cluster with surfaces of complex shape and non-linear, allowing a better interpretation of the results... [edited by author]
Description: 2009 - 2010
URI: http://hdl.handle.net/10556/1439
http://dx.doi.org/10.14273/unisa-284
Appears in Collections:Economia del settore pubblico

Files in This Item:
File Description SizeFormat 
tesi_A_Ciani.pdftesi di dottorato3 MBAdobe PDFView/Open
abstract_in_italiano_e_in_inglese_A_Ciani.pdfabstract in italiano a cura dell'autore109,25 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.