An Implicit Expectation Conditional Maximization Algorithm for Non-homogeneous Poisson Process Software Reliability Models
Vidhyashree Nagaraju
Uniiversity of Massachusetts
Lance Fiondella
University of Massachusetts Dartmouth
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 NHPP SRGM. In contrast to the expectation maximization algorithm, 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, enabling implicit solutions to update rules when a close form equation is not available for a model parameter. Two variants are proposed.