ABSTRACT
Quality, cost, and delivery are important software development management factors that software development managers control during the testing phase of the software development process. In particular, the software development management techniques based on software reliability growth models (SRGM) are often used. The majority of SRGMs are developed by treating the software failure-occurrence time or the fault-detection time intervals as random variables, and they assume that the stochastic characteristics of these quantities are the same throughout the testing phase. However, researchers often observe that stochastic characteristics of software failure-occurrence time intervals change because of a change in the testing environment. The chapter discusses the optimal software release problem based on the change-point model as an application. It describes the optimal software release problem based on the multi-attribute utility theory (MAUT). The MAUT is a utility theory on decision-making considering multiple constraints.
