Parameter Estimation of Software Reliability Growth Model

Abstract

Reliability normally covers each and every part of the device thinking about hardware, software program and processes but software program reliability paperwork an important factor as we are continually worried with efficient and quality software program. Software reliability is a crucial characteristic of software, nice for predicting the diploma of creditability of the precise software for the specified time frame and in particular surroundings. Software reliability models are categorized into two; one is the static version in which modeling and evaluation of software logic is performed at the same code, different one is dynamic model that observes the brief behavior of debugging technique in the course of checking out section. But the model parameters are in nonlinear relationships which create many troubles in locating parameters which might be best using traditional techniques like maximum likelihood and least rectangular estimation. Various algorithms have been introduced which makes the task of parameter estimation less complicated. Parameter estimation of Non-homogenous Poisson Procedure (NHPP) fashions, the use of MLE (Maximum likelihood Estimation) and seek algorithms referred to as particle swarm optimization has been explored in this paper. The incorporation of different device learning procedures within the prediction of software program reliability has proven enhancements in comparison to the statistical strategies.

Keywords: Software reliability, Support vector machine Software reliability growth models, Maximum likelihood Estimation, Particle Swarm Intelligence, parameter estimation

Cite this Article Bisma Gulzar Mir, Sheikh Riyaz-ul-Haq, Parameter Estimation of Software Reliability Growth Model. Journal of Artificial Intelligence Research & Advances. 2020; 7(1): 15–22p.