In this paper, two models predicting mean time until next failure based on
Bayesian approach are presented. Times between failures follow Weibull dist
ributions with stochastically decreasing ordering on the hazard functions o
f successive failure time intervals, reflecting the tester's intent to impr
ove the software quality with each corrective action. We apply the proposed
models to actual software failure data and show they give better results u
nder sum of square errors criteria as compared to previous Bayesian models
and other existing times between failures models, Finally, we utilize likel
ihood ratios criterion to compare new models predictive performance.