Application of Random forest method to estimate the incurred but not reported claims reserve of an insurance company
Abstract
The purpose of this report is to explore the applicability of Random forest method to assess the incurred but not reported claims reserve (IBNR) of a non-life insurance company. The research is based on the statistical method of Random forest. The actual data on the direct hull insurance of two real companies for the period 2009-2014 were used. The IBNR valuated on 31.12.2014 by Random forest was compared with the results of standard calculation methods (chain ladder and Bornhuetter − Ferguson on paid triangles). In general, we can say that the Random forest method can be applied to assess the IBNR as an alternative algorithm.
Full Text:
PDF (Russian)References
Breiman L. Random forests // Machine learning. 2001. 45(1). R. 5-32.
Siroky D. Navigating Random Forests and related advances in algorithmic modeling // Statistics Surveys, 3. 2009. P. 147–163.
Breiman L., Friedman R., Olshen R., & Stone C. Classification and Regression Trees // Belmont, California: Wadsworth International. 1984.
Breiman L. Out-of-bag estimation // Berkeley: Technical Report, Statistics Department University of California. 1996. 13 r.
Liaw A., Wiener M. Classification and Regression by Random Forest // R News, 2 (3). 2002. R. 18-22.
Chistjakov S. Sluchajnye lesa: obzor // Trudy Karel'skogo nauchnogo centra RAN (1). 2013. S. 117-136.
Gruzdev A.V. Metod sluchajnogo lesa v skoringe // Risk-menedzhment v kreditnoj organizacii (#1 (13)). 2014. S. 28-43.
Breiman L. Manual on setting up, using, and understanding random forests v 4.0. URL (data obrashhenija 22.10.2015 g.):
https://www.stat.berkeley.edu/~breiman/papers.html
Refbacks
- There are currently no refbacks.
Abava Кибербезопасность IT Congress 2024
ISSN: 2307-8162