A Comparison of Classification Models for Life Insurance Lapse Risk
Lim Jin Xong1, Ho Ming Kang2
1Lim Jin Xong, School of Mathematics, Actuarial and Quantitative Studies, Asia Pacific University of Technology and Innovation, Malaysia.
2Ho Ming Kang, School of Mathematics, Actuarial and Quantitative Studies, Asia Pacific University of Technology and Innovation, Malaysia.
Manuscript received on 05 February 2019 | Revised Manuscript received on 11 February 2019 | Manuscript Published on 19 February 2019 | PP: 245-250 | Volume-7 Issue-5S January 2019 | Retrieval Number: ES2151017519/19©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Insurer usually incurs expenses such as policy issuance cost, commission and administrative costs after the launching of an insurance product. When a policyholder decided to lapse a policy, the insurer have to seek for alternative to cover all the losses, which might liquidate high-yielding investments in order to satisfy their requests for the cash value or surrender value. Therefore, it is crucial to understand and develop a classification model to determine the surrender or lapse risk. In this study, four classification models such as logistic regression, k-Nearest Neighbor, Neural Network (NN) and Support Vector Machines (SVM) are used to model the life insurance lapse risk, which is the risk thatinvolving the termination of policies by the policyholders. Classification performance criterions such as prediction accuracy and area under the Receiver Operating Curve (ROC) are used to compare the performance between the models. The results showed that SVMwas outperformed than NN, logistic regression and k- Nearest Neighbor.
Keywords: Logistic Regression, k-Nearest Neighbor, Neural Network, Support Vector Machines, Classification, Lapse Risk.
Scope of the Article: Classification