Table 3.

Comparison of algorithms shortcomings with PC-SVM model

S. No	Algorithms	Shortcoming
1	AdaBoost	This algorithm is affected by outliers and not effective in predicting the errors but SVM is free from outliers and effective in predicting the software defects [7]
2	CART	It has poor modeling with linear data while SVM can work with both linear as well as non-linear datasets [7]
3	KNN	It has a high rate of classification in comparison with SVM [7]
4	Neural Network	It works good with large datasets and taking time in training the datasets while our model has small datasets for that PC-SVM is better [7]
5	Chao Genetic	It has difficulty in providing the optimal solution while PC can provide [7]
6	EM model	It also does not guarantee in providing the optimal solution but our model PC-SVM can give [7]