Software Engineering is seamlessly developed with advancement in the development tools and framework as so the development of the defect also. Software defect prediction focus on ensuring and delivering quality software product. Machine learning algorithms plays a vital role in effectively predicting the software bugs early in the software life cycle. The main objective of this study is to identify a hybrid machine learning model incorporating Correlation feature selection. The radial basis function kernel of SVM maps the features in the input space. The CFS based SVM model encompasses a higher defect prediction when employed. The CM1 dataset from NASA spacecraft instrument were subjective for the evaluation of CFS based SVM model. Confusion matrix, Precision, Recall, and F1 Score are the metrics for CFS based SVM. The results show a higher rate of prediction with 64.32% for precision, 65.13% for Recall, and 64.72% for F Score, respectively. This study has been able to show the defective classes have a big impact on software defect prediction and opting the correct parameters with correlation feature selection effectively performs predictions among the other classifiers.