Developing Burr-XII NHPP-based software reliability growth model using Expectation Conditional Maximization Algorithm
Sheng Han, Qiang Han, Yixin Qiao, Kehan Xue, Zhichao Shi · 2024
With the increasing complexity of software systems, the development of models that can accurately predict and enhance software reliability has become particularly important. This paper introduces an innovative method, applying the Expectation Conditional Maximization (ECM) algorithm to the Burr-XII Software Reliability Growth Model (SRGM) based on the Non-Homogeneous Poisson Process (NHPP). The ECM algorithm, as a variant of the Expectation Maximization (EM) algorithm, is particularly suitable for the estimation of complex model parameters. We utilize the ECM algorithm to estimate the parameters of the Burr XII NHPP SRGM and compare its performance with traditional parameter estimation methods. The results of the experiment demonstrate superior goodness-of-fit and predictive ability on multiple datasets compared to traditional methods, which highlights the efficiency and accuracy of the ECM algorithm in parameter estimation.