Abstract:To facilitate developers in effective allocation of their testing and debugging efforts, many software defect prediction techniques have been proposed in the literature. These techniques can be used to predict classes that are more likely to be buggy based on the past history of classes, methods, or certain other code elements. These techniques are effective provided that a sufficient amount of data is available to train a prediction model. However, sufficient training data are rarely available for new software projects. To resolve this problem, cross-project defect prediction, which transfers a prediction model trained using data from one project to another, was proposed and is regarded as a new challenge in the area of defect prediction. Thus far, only a few cross-project defect prediction techniques have been proposed. To advance the state of the art, in this study, we investigated seven composite algorithms that integrate multiple machine learning classifiers to improve cross-project defect prediction. To evaluate the performance of the composite algorithms, we performed experiments on 10 open-source software systems from the PROMISE repository, which contain a total of 5,305 instances labeled as defective or clean. We compared the composite algorithms with the combined defect predictor where logistic regression is used as the meta classification algorithm (CODEP (Logistic) ), which is the most recent cross-project defect prediction algorithm in terms of two standard evaluation metrics: cost effectiveness and F-measure. Our experimental results show that several algorithms outperform CODEP (Logistic) : Maximum voting shows the best performance in terms of F-measure and its average F-measure is superior to that of CODEP (Logistic) by 36.88%. Bootstrap aggregation (Bagging (J48)) shows the best performance in terms of cost effectiveness and its average cost effectiveness is superior to that of CODEP (Logistic) by 15.34%.

Predicting Upgrade Project Defects Based on Enhancement Requirements: An Empirical Study

Towards More Practical Automation of Vulnerability Assessment

Multi-project Regression Based Approach for Software Defect Number Prediction

Combined Classifier for Cross-Project Defect Prediction: an Extended Empirical Study.

Connecting Software Metrics across Versions to Predict Defects

Software-defect prediction within and across projects based on improved self-organizing data mining

Explainable Software Defect Prediction from Cross Company Project Metrics Using Machine Learning

An Improved Software Reliability Prediction Model by Using High Precision Error Iterative Analysis Method.

A Study of Software Defect Prediction Using Artificial Intelligent and Business Risk Management Approach

Perceptions, Expectations, and Challenges in Defect Prediction.

Towards Developing and Analysing Metric-Based Software Defect Severity Prediction Model

A Novel Evaluation Method for Defect Prediction in Software Systems

Quantitative Analysis of Requirements Evolution across Multiple Versions of an Industrial Software Product

Revisiting Unsupervised Learning for Defect Prediction

The Impact of Defect (Re) Prediction on Software Testing

Enhancing and Measuring the Predictive Capabilities of Testing-Effort Dependent Software Reliability Models

FEPP: Advancing Software Risk Prediction in Requirements Engineering Through Innovative Rule Extraction and Multi-Class Integration

An Empirical Study on Heterogeneous Defect Prediction Approaches

Improving Software Defect Prediction by Combining the Information of Class and Package

A Software Defect Prediction Approach Based on Hybrid Feature Dimensionality Reduction

Towards Cross-Project Defect Prediction with Imbalanced Feature Sets