软件可靠性预测模型的优选问题研究

The Study on the Optimization of Software Reliability Prediction Model

针对软件可靠性预测模型过多且难以选择的实际问题,基于混合专家网络,采用基于梯度下降法的变步长LMS算法,解决软件可靠性预测模型选择优化问题.并通过实例分析,验证了该算法应用于模型优选的有效性.

This paper is to solve the problems that software reliability prediction models are excessive and difficult to select, based on mixed network of experts, based on the gradient descent method variable step size LMS algorithm to solve software reliability prediction model selection optimization problem. Finally, the algorithm is applied to verify the model optimal effectiveness.