LNCS Homepage
CD ContentsAuthor IndexSearch

Learning Environment for Life Time Value Calculation of Customers in Insurance Domain

Andrea Tettamanzi1, Luca Sammartino2, Mikhail Simonov2, Massimo Soroldoni2, and Mauro Beretta3

1Università degli Studi di Milano, Dipartimento di Tecnologie dell’Informazione, Via Bramante 65, I-26013 Crema (CR), Italy
andrea.tettamanzi@unimi.it

2Nomos Sistema S.p.A., Viale Monza 259, I-20126 Milano (MI), Italy
sammartino@nomos.it
simonov@nomos.it
soroldoni@nomos.it

3Genetica S.r.l., Via San Dionigi 15, I-20139 Milano (MI); Italy
beretta@genetica-soft.com

Abstract. A critical success factor in Insurance business is its ability to use information sources and contained knowledge in the most effective way. Its profitability is obtained through the Technical management plus Financial management of the funds gathered on the market. The profitability of a given customer can be evaluated through its Life Time Value (LTV). We aim at applying evolutionary algorithms to the problem of forecasting the future LTV in the Insurance Business. The Framework developed within the Eureka cofunded research projects HPPC/SEA and IKF has been adapted to the Insurance Domain through a dedicated Genetic Engine. The solution uses RDF and XMLcompliant standard. The idea of using evolutionary algorithms to design fuzzy systems date from the beginning of the Nineties and a fair body of work has been carried out throughout the past decade. The approach we followed uses an evolutionary algorithm to evolve fuzzy classifiers of the data set.

LNCS 3103, p. 1251 ff.

Full article in PDF


lncs@springer.de
© Springer-Verlag Berlin Heidelberg 2004