A Hybrid Support Vector Machine–Genetic Algorithm Framework Approach for Parameters Estimation for the [0,1] Truncated Nadarajah Haghighi Inverse Rayleigh Process in nonhomogenous Poisson Process

Abstract

The estimation of three unknown parameters of [0,1] Truncated Nadarajah Haghighi Inverse Rayleigh Process of nonhomogeneous Poisson Process (TNHIRP). It discusses nine of those estimation methods, which are the maximum likelihood, maximum product spacing, Anderson Darling, Right Anderson-Darling, ordinary least squares, weighted least squares, Cramer-von Mises, Hybrid Support Vector Machine–Genetic Algorithm, and percentiles to estimate points. These estimators were compared in terms of bias and mean squared errors using detailed analysis and a lot of simulation experiments. According to the simulation outcomes, all the estimators work well under these considerations and show similar values in case of large sample size. There are also some statistical properties of the new distribution derived. The estimators were also applied to two real life datasets and the values of the Kolmogorov-Smirnoff statistics reported. Lastly, the [0,1] Truncated Nadarajah Haghighi Inverse Rayleigh process was tested on two real world datasets, and the performance was compared with other well-known extensions of Inverse Rayleigh process using four selection criteria.