Authors:
Athanasios Tsakonas
and
Georgios Dounias
Affiliation:
University of the Aegean, Greece
Keyword(s):
Genetic programming, software engineering, data mining, effort estimation, defect prediction.
Related
Ontology
Subjects/Areas/Topics:
Applications of Expert Systems
;
Artificial Intelligence
;
Artificial Intelligence and Decision Support Systems
;
Computational Intelligence
;
Enterprise Information Systems
;
Evolutionary Computing
;
Knowledge Acquisition
;
Knowledge Discovery and Information Retrieval
;
Knowledge Engineering and Ontology Development
;
Knowledge-Based Systems
;
Machine Learning
;
Soft Computing
;
Symbolic Systems
Abstract:
Research in software engineering data analysis has only recently incorporated computational intelligence methodologies. Among these approaches, genetic programming retains a remarkable position, facilitating symbolic regression tasks. In this paper, we demonstrate the effectiveness of the genetic programming paradigm, in two major software engineering duties, effort estimation and defect prediction. We examine data domains from both the commercial and the scientific sector, for each task. The proposed model is proved superior to past literature works.