A Prediction Model Based on ISOMAP for Software Defects

Abstract:

To improve and guarantee the quality of software, it is very necessary to effectively predicting modules with defects in the software. There are usually more measure attributes in software quality prediction, which often leads to the curse of dimension. To do this, a new algorithm based on ISOMAP was presented to predict software defect, which combined manifold learning algorithms and classification methods. In the model, the high dimensional software metrics attribute data were firstly mapped into the low dimensional space through ISOMAP. Then the low dimensional features were classified with KNN, SVM and NB. Experiments demonstrate that the new model progresses the prediction precision of software defects as well as great improves the efficiency of the algorithm.

• Periodical: Applied Mechanics and Materials (Volumes 347-350)
• Pages: 3278-3282
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.347-350.3278
• Online since: August 2013
• Authors: Sheng Li Shi, Jin Shi, Rui Wang
• Keywords: Isomap, KNN, Software Defect Prediction, SVN

References

[1] N. E. Fenton, N. Ohlsson, Quantitative analysis of faults and failures in a complex software system, IEEE Transactions on Software Engineering, 2000, vol. 26, no. 8, p.797–814.

DOI: 10.1109/32.879815

[2] A. Koru, J. Tian, An empirical comparison and characterization of high defect and high complexity modules, Journal of Systems and Software, 2003, vol. 67, no. 3 , p.153–163.

DOI: 10.1016/s0164-1212(02)00126-7

[3] A. Koru, H. Liu, Building effective defect-prediction models in practice, IEEE Software, 2005, vol. 22, no. 7, p.23–29.

DOI: 10.1109/ms.2005.149

[4] T. M. Khoshgoftaar, N. Seliya, Tree-based software quality estimation models for fault prediction, Proceedings of the 8th IEEE Symposium on Software Metrics, 2002, Ottawa, Canada, pp.203-214.

DOI: 10.1109/metric.2002.1011339

[5] Mie Mie Thet Thwin, Tong-Seng Quah, Application of neural networks for software quality prediction using object-oriented metrics, Journal of Systems and Software, 2005, vol. 76, no. 2, p.147–156.

DOI: 10.1109/icsm.2003.1235412

[6] T. Menzies, J. Greenwald, A. Frank, Data mining static code attributes to learn defect predictors, IEEE Transactions on Software Engineering, 2007, vol. 33, no. 1, pp.2-13.

DOI: 10.1109/tse.2007.256941

[7] K. O. Elish, M. O. Elish, Predicting defect-prone software modules using support vector machine, Journal of Systems and Software, 2008, vol. 81, no. 5, pp.649-660.

DOI: 10.1016/j.jss.2007.07.040

[8] H. Jiang, M. Zong, X. Liu, Research of software defect predication model based on ACO-SVM, Chinese Journal of Computers, 2011, vol. 34, no. 6, pp.1149-1154.

DOI: 10.3724/sp.j.1016.2011.01148

[9] J. B. Tenenbaum, V. de Silva, J. C. Langford, A global geometric framework for nonlinear dimensionality reduction, Science, 2000, vol. 290, pp.2319-2323.

DOI: 10.1126/science.290.5500.2319

[10] S. T. Roweis, L. K. Saul, Nonlinear dimensionality reduetion by loeally linear embedding. Scienee, 2000, vol. 290, pp.2323-2326.

DOI: 10.1126/science.290.5500.2323

[11] M. Belkin, P. Niyogi, Laplacian eigenmaps for eimensionality rReduction and eata representation, Neural Computation, 2003, vol. 15, no. 6, pp.1373-1396.

DOI: 10.1162/089976603321780317

[12] Z. Zhang, H. Zha, Principal manifolds and nonlinear dimensionality reduction via tangent space alignment, SIAM Journal of Scientific Computing, 2004, vol. 26, no. 1, pp.313-338.

DOI: 10.1137/s1064827502419154

[13] http: /mdp. ivv. nasa. gov/index. html.