Downloads: 119
Research Paper | Statistics | Volume 5 Issue 4, April 2016 | Pages: 2025 - 2028 | India
Kernel Density Estimation for Claim Size Distributions Using Shifted Power Transformation
Abstract: This paper presents density estimation of univariate claim severity distributions using kernel density estimation. We applied transformations to data prior to implement kernel density estimation so as to ensure the data is symmetry for the purpose of applying Gaussian methods. The paper presents non-parametric method of obtaining density for univariate claim severity distributions with goodness fits analysis for Danish insurance data on fire loss claims.
Keywords: Kernel density, bandwidth, Power transformation, Loss modeling, Cross validation
How to Cite?: K. M. Sakthivel, C. S. Rajitha, "Kernel Density Estimation for Claim Size Distributions Using Shifted Power Transformation", Volume 5 Issue 4, April 2016, International Journal of Science and Research (IJSR), Pages: 2025-2028, https://www.ijsr.net/getabstract.php?paperid=NOV163071, DOI: https://dx.doi.org/10.21275/NOV163071