Expectation Conditional Maximization Algorithms for Failure count Non-homogeneous Poisson Process Software Reliability Models  
Author Vidhyashree Nagaraju

 
Co-Author(s) Anusha Krishna Murthy; Lance Fiondella; Panlop Zeephongsekul; Thierry Wandji

 
Abstract Software reliability growth models (SRGM) based on the non-homogeneous Poisson process (NHPP) are a popular approach to estimate useful metrics such as the number of faults remaining, failure rate, and reliability. However, it is often difficult to apply SRGM in practice because even relatively simple models can require numerical solution of complex systems of equations. To overcome this limitation, we propose expectation conditional maximization (ECM) algorithms for failure count NHPP SRGM. The ECM algorithm reduces the maximum likelihood estimation process to multiple simpler conditional maximization (CM)-steps. The advantage of these CM-steps is that they only need to consider one variable at a time. We illustrate the ECM algorithm by applying it to a failure count data set from the research literature. Our results demonstrate that the algorithm identifies the MLEs, monotonically improving the log-likelihood function in each iteration.

 
Keywords Software reliability, expectation conditional maximization algorithm, failure count data
   
    Article #:  22372
 
Proceedings of the 22nd ISSAT International Conference on Reliability and Quality in Design
August 4-6, 2016 - Los Angeles, California, U.S.A.