AN EMPIRICAL APPROACH FOR COMPLEXITY REDUCTION AND FAULT PREDICTION FOR SOFTWARE QUALITY ATTRIBUTE
S. Duraisamy, N. Rajkumar, N.A. Viji · International Journal of Business Intelligence and Data Mining · 2017
Designing the high-quality software is a difficult one due to the high complexity and fault prone class. To reduce the complexity and predict the fault-prone class in the object orient software design proposed a new empirical approach. This proposed approach concentrates more on to increase the software quality in the object oriented programming structures. This technique will collect the dataset and metric values from CK-based metrics. And then complexity will be calculated based on the weighted approach. The fault prediction will be done, based on the low usage of the dataset and high complexity dataset. This helps to increase the software quality. In simulation section, the proposed approach has performed and analysed the parameters such as accuracy, fairness, recall, prediction rate and efficiency. The experimental results have shown that the proposed approach increases the prediction rate, accuracy and efficiency.