基于神经网络的软件质量综合评价方法

An Overall Evaluating Method for Software Qualities Based on Neural Networks

摘要: 研究了软件质量评价的度量模型，指出了软件综合质量评价中需要解决的一致性与稳定性问题，提出了基于前向单隐层神经网络的软件质量评价方法，一个软件质量评价实例表明，该方法具有满意的效果。

Abstract: This paper studies metric models of software quality evaluations and in particular looks into consistency and stability in global quality evaluations. Furthermore, this paper proposes a global evaluation method based on feedforward single-hidden-layer neural networks. An application result shows that the presented method is quite effective.