Title: A high-order feature synthesis and selection algorithm applied to insurance risk modelling

Authors: Charles Dugas, Nicolas Chapados, Rejean Ducharme, Xavier Saint-Mleux, Pascal Vincent

Addresses: ApSTAT Technologies Inc., 4200, boul. St-Laurent, suite 408, Montreal, Quebec, H2W 2R2, Canada. ' ApSTAT Technologies Inc., 4200, boul. St-Laurent, suite 408, Montreal, Quebec, H2W 2R2, Canada. ' ApSTAT Technologies Inc., 4200, boul. St-Laurent, suite 408, Montreal, Quebec, H2W 2R2, Canada. ' ApSTAT Technologies Inc., 4200, boul. St-Laurent, suite 408, Montreal, Quebec, H2W 2R2, Canada. ' ApSTAT Technologies Inc., 4200, boul. St-Laurent, suite 408, Montreal, Quebec, H2W 2R2, Canada; Universite de Montreal, Pavillon Andre-Aisenstadt 2920 chemin de le Tour, local 2194, Montreal, Quebec, H3T 1J4, Canada

Abstract: In many jurisdictions, automobile insurers have access to risk-sharing pools to which they can transfer some risks. We consider different feature selection and modelling approaches to maximise profitability of these transfers through better risk selection. For that purpose, we introduce a flexible scoring model and devise a robust feature synthesis and selection method. We show what should be the most suitable sorting criterion depending on pool regulations. We use a technique, similar to cross validation, but that is coherent with the sequential structure of insurance data. We explain how software maturity level impacts profitability.

Keywords: business intelligence; feature extraction; feature synthesis; feature selection; scoring models; bagging; sequential cross validation; profit maximisation; automotive insurance; automobile industry; risk-sharing pools; loss ratios; modelling; software maturity level; car insurance.

DOI: 10.1504/IJBIDM.2011.041957

International Journal of Business Intelligence and Data Mining, 2011 Vol.6 No.3, pp.237 - 258

Published online: 22 Apr 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article