An Early Prediction Method of Software Reliability Based on Support Vector Machine

With rapid development of information technology, we are more and more dependent on computer systems. Computer software is the primary factor among all that can result in computer system errors. Therefore, how to design and develop reliable software, and how to ensure its quality is an urgent task. Early prediction of software reliability can enable developers to obtain general ideas about software reliability before testing it. This is critical for further development, testing and quality control of the software. This paper summarizes the status of early prediction methods for software reliability in both China and in other countries. We have introduced the support vector machine (SVM) theory into the study on early prediction of software reliability, and advanced the software reliability early prediction model based on SVM. Simulation shows that comparing with the classic models, this model is more precise in its prediction capacity, and has better capacity in generalization, and is less dependent on the sample size.