We discuss a simulation-based interval estimation method of software reliability and cost-optimal software release time based on a nonparametric bootstrap method via a discretized software reliability growth model. In general, it is very difficult to derive probability distributions of model parameters, software reliability assessment measures, and optimal software release time analytically even if we assume an asymptotic theory assuming a large number of samples. The bootstrap method overcomes the problem above by obtaining the probability distributions of the parameter based on the simulation-based approach. This paper includes numerical examples for confirming that our bootstrap approach yields simulation-based probability distributions of model parameters, software reliability assessment measures, and cost-optimal software release time from actual software fault-count data.
Interval Estimation Method for Decision Making in Wavelet-Based Software Reliability Assessment
