Principal Component Analysis' Application to the Software Metrics-based for Risk Assessment
Hao Wang · Yunchou yu guanli · 2005
Software risk assessment techniques can be divided into subjective methods and objective methods.The subjective methods are often dependent on the knowledge of experts and the experiment of managers,while the objective methods are based on the metrics of software products' attributes such as complexity.In this paper,the principal component analysis,which belongs to the modern statistic analytical techniques,is used to assess the software risk.A case study for using PCA to assess risk of the object-oriented software development is also given.In doing so,the software developers or management people can identify the software high-risk modules easily and perform risk management effectively.