基于随机向量镜像策略改进ART算法
发布时间：2018-11-27 07:42
【摘要】：针对现有的镜像自适应随机测试(MART)、动态镜像自适应随机测试(DMART)等算法通过镜像函数生成的测试用例的随机性不足,使其有效性在不同程度上有明显下降的问题,提出基于随机向量镜像策略改进ART算法.首先将随机向量引入传统镜像函数,增大镜像测试用例间的差异性;然后将随机向量镜像函数运用到镜像策略中,改进现有的ART算法.实验结果表明,利用随机向量镜像策略可明显地提高镜像算法的有效性,并且该算法比传统ART算法的效率有显著提升.
[Abstract]:Aiming at the lack of randomness of the test cases generated by the existing image adaptive random test (MART),) dynamic mirror adaptive random test (DMART) algorithms by mirror function, the effectiveness of these algorithms is obviously reduced in different degrees. An improved ART algorithm based on random vector mirroring strategy is proposed. Firstly, the random vector is introduced into the traditional mirror function to increase the difference between the mirror test cases, and then the random vector image function is applied to the mirror strategy to improve the existing ART algorithm. The experimental results show that the effectiveness of the image algorithm can be improved obviously by using the random vector image strategy, and the efficiency of this algorithm is significantly improved than that of the traditional ART algorithm.
【作者单位】： 解放军信息工程大学;数学工程与先进计算国家重点实验室;
【基金】：国家自然科学基金(61402525) 郑州市普通科技攻关项目(141PPTGG383)
【分类号】：TP301.6;TP311.53
本文编号：2359910
[Abstract]:Aiming at the lack of randomness of the test cases generated by the existing image adaptive random test (MART),) dynamic mirror adaptive random test (DMART) algorithms by mirror function, the effectiveness of these algorithms is obviously reduced in different degrees. An improved ART algorithm based on random vector mirroring strategy is proposed. Firstly, the random vector is introduced into the traditional mirror function to increase the difference between the mirror test cases, and then the random vector image function is applied to the mirror strategy to improve the existing ART algorithm. The experimental results show that the effectiveness of the image algorithm can be improved obviously by using the random vector image strategy, and the efficiency of this algorithm is significantly improved than that of the traditional ART algorithm.
【作者单位】： 解放军信息工程大学;数学工程与先进计算国家重点实验室;
【基金】：国家自然科学基金(61402525) 郑州市普通科技攻关项目(141PPTGG383)
【分类号】：TP301.6;TP311.53
【相似文献】
相关期刊论文 前6条
1 宗峰;;实例推理中的ART网络应用[J];电脑知识与技术;2009年17期
2 林桦，王正光;ART网在地图彩色分类中的应用[J];西安电子科技大学学报;1994年04期
3 刘芳;晏志勇;;基于ART的电子商务推荐系统的研究与实现[J];中国商贸;2009年07期
4 许令周;范宜仁;夏卫芳;;ART算法中射线矩阵加权的BPT算法[J];计算物理;2011年05期
5 赵祥模;李娜;关可;宋焕生;;基于LTI射线追踪的改进的ART算法[J];中国图象图形学报;2009年05期
6 常振兴;杨立瑞;;多视角重建ART算法优化研究[J];警察技术;2008年04期
，本文编号：2359910
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2359910.html
