A comparison of techniques for developing predictive models of software metrics

摘要: The use of regression analysis to derive predictive equations for software metrics has recently been complemented by increasing numbers studies using non-traditional methods, such as neural networks, fuzzy logic models, case-based reasoning systems, and trees. There also an level sophistication in the regression-based techniques used, including robust factor analysis, more effective validation procedures. This paper examines implications these methods provides some recommendations when they may be appropriate. A comparison various is made terms their modelling capabilities with specific reference metrics.