The problem of software fault content and reliability estimations is considered. Estimations that can be used to improve the control of a software projects are emphasized. A model of how to estimate the number of software faults is presented, which takes into account both the development process and the developed software product. A model of how to predict, before the testing has started, the occurrences of faults during the testing stage is also presented. These models, together with software reliability growth models, have been evaluated on a number of software projects.
展开▼
