Component Consistency during Design: A Rule based Evaluation Technique
N. Rajasekhar Reddy, K.V.S. Praveena · 2012
The automated detection of faulty modules contained by software systems might lead to reduced development expenses and additional reliable software. In this effort design and development metrics has been used as features to predict defects in given software module using SVM classifier. A rigorous progression of pre-processing steps were applied to the data preceding to categorization, including the complementary of in cooperation classes (faulty or otherwise) and the elimination of a large numeral of repeating instances. The Support Vector Machine in this trial yields a standard accuracy that miles ahead over existing defect prediction models on previously unseen data.