HWANG-HA-JIN-IDA2009-Abstract(Athens)

advertisement
Applying Genetic Algorithm to Integrate Data Mining Models for
Customer Financial Status Evaluation
Ha Jin Hwang
Department of Management Information Systems, Catholic University of Daegu,
Kyungsan-si, Kyungbuk, Korea
hjhwang@cu.ac.kr
Abstract. Customer financial status evaluation has been regarded as one of the most important tools by financial
institutions to assess the ability of prospective customers to handle their financial matters and to monitor customer
credit ratings. Based on details of financing and payment histories from a financial institution, customer financial
status evaluation can be conducted. In this study, individual data mining models using artificial neural network, MDA,
and decision tree analysis were derived. The results obtained from these single models were subsequently compared
with the results from an integrated model that was developed using genetic algorithm. This study not only verifies
existing individual models and but also attempts to overcome the limitations of these approaches.
While the
comparative analysis of individual models for the purpose of identifying the best-fit model relies upon existing
techniques, this study addresses a new approach to develop an integrated data mining model using genetic algorithm.
Keywords: data mining, genetic algorithm, customer financial status evaluation
Download