%0 Journal Article
%T 基于GAM模型对失效数变化的分析<br>Analysis of Failure Number Change Based on GAM Model
%A 杨玉鑫
%J Computer Science and Application
%P 1255-1265
%@ 2161-881X
%D 2019
%I Hans Publishing
%R 10.12677/CSA.2019.97141
%X <div style=\"text-align:justify;\">
	<span style=\"font-size:14px;\">本文针对软件生产进程中测试的过程数据，利用带有交互作用的GAM，对测试过程中的失效变化数据进行分析，研究影响因素以及因素间的交互作用。本文首先对失效变化数据，比较了GAM和带有交互作用的GAM的拟合效果；其次，比较了带有交互作用的GAM与现有的机器学习模型的预测结果。结果表明，带有交互作用的GAM相较于GAM所得的结果更加丰富，且优于已有的机器学习方法，因此这样的探讨分析是有意义的，并对软件开发以及采取测试决策有一定的实际指导意义。<br />
In this paper, the process data in the process of software production is used to analyze the failure change data in the test process by using GAM with interaction, and to study the influencing factors and the interaction between the factors. In this paper, the fitting effects of GAM and GAM with interaction are compared firstly with the failure change data. Secondly, the prediction results of GAM with interaction and existing machine learning model are compared. The results show that the GAM with interaction is more abundant than that obtained by GAM, and it is better than the existing machine learning method. Therefore, such analysis and analysis is meaningful.</span> 
</div>
%K 失效数，GAM模型，交互作用<br>Failure Number
%K GAM Model
%K Interaction
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=31294