MACHINE LEARNING - TECHNIQUES

Abstract

This article provides a comprehensive overview of software development expertise using machine learning techniques (MLT). Machine learning in this new era demonstrates the commitment to consistently make accurate estimates. The machine learning system effectively “learns” how to evaluate from the training package of completed projects. The main goal and contribution of the review is to support research on expert assessment, i.e. to facilitate other researchers to make relevant expert assessment studies using machine learning techniques. This article presents commonly used machine learning techniques such as neural networks for expert evaluation in the field of software development, case-based reasoning, classification and regression trees, induction, genetic algorithm and genetic programming. In each of our studies, we found that the results of different machine learning techniques depend on the areas in which they are used. The review of our study not only indicates that these techniques compete with traditional evaluators in a data set, but also illustrate that these methods are sensitive to the data on which they are trained.