Predicting Time Series Data collected from Software Measurements with Machine Learning Approaches

Abstract

The objective of this paper is to highlight the implementation of machine learning forecasting approaches in software development. The concept of data mining has been used in different areas in the industry. There is an existing gap in the field of applying machine learning in the context of software measurements. This thesis will be conducted in two parts. Part 1, a systematic literature review to pinpoint the most recognised machine learning approaches. While part 2 will test the found approaches in an experimental environment to determine the most suitable machine learning approach for the collected data. The data was collected in a previous study through a collection of automotive software measurements.