Abstract
Reliability demonstration testing for software
products is performed for the purpose of examining whether
the specified reliability is realized in the software after the
development process is completed. This study proposes a model
of reliability demonstration testing for discrete-type software
such as software for numerical calculations. The number of
input data sets for test and acceptance number of input data
sets causing software failures in the test are designed based
on variation distance. This model has less parameters to be
prespecified than the statistical model.
Users
Please
log in to take part in the discussion (add own reviews or comments).