Machine Learning Applications in Software Engineering

In this research project, we apply machine learning techniques to problems from software engineering domain. This project is a collaboration with the Theoretical Computer Science Group, espcially Prof. Waack. As part of the current research, Steffen Herbold has written his masters thesis with the title "Detection of feature freezes using clustering algorithms", which was awared with the DASMA-Diplomarbeitenpreis 2008. Additionally, we investigated how machine learning can be applied to calculate thresholds for software metric sets in order to optimize them.Another direction of research is the reconstruction of system specifications from test traces. To this aim, Edith Werner applied automaton learning to test traces of a system to calculate an automaton that specifies the system.