Overview

Depending on the domain from which the training data are collected and the nature of the target function to be learned, ML methods can be used to generate or synthesize various types of software products or artifacts. This chapter pertains to the applications of ML methods for software artifacts generation or synthesis. Examples include: test data generation, synthesis of test-resource allocation, generation of project management rules or schedules, and synthesis of agent programs, data structures, scripts, or design schemas. Table 25 provides a glimpse of the ML methods used in this application area.