基于广义分形插值理论的多尺度分类尺度下推算法
|
摘要
多尺度数据挖掘多应用于空间遥感图像数据,以图像的分辨率或者区域分割为依据进行尺度划分,然后在每个尺度层进行分析。近期,有不少学者将多尺度数据挖掘应用于一般数据集上,以等级理论、概念分层以及包含度理论等为尺度划分依据,研究不同尺度层的分布规律,进而发现有意义的事实,如多尺度关联规则以及多尺度聚类。但是在一般数据集下很少将多尺度数据挖掘应用于分类算法领域。定义了广义分形插值理论的概念,打破了局限于迭代函数系统(iterative function systems,IFS)的缺憾,拓展了分形插值的应用;提出了基于广义分形插值理论的多尺度分类尺度下推算法(multi-scale classification scaling-down algorithm,MSCSDA)。仿真实验建立在四个UCI基准数据集和一个H省部分人口真实数据集上,并将MSCSDA与KNN、decision tree以及LIBSVM算法进行对比分析,实验结果表明,MSCSDA在不同的数据集上均优于其他算法。
The research of multi-scale data mining mainly applies to space remote sensing image data sets,and conducts scale division based on the resolution or regional segmentation of the images,then analysis knowledge on each scale layer. Recently,there are quite a few learners apply the multi-scale data mining to general data sets,and conduct scale division based on the level theory,concept hierarchy and inclusion degree etc.,study the distribution rule on different scale layers,and then found significant facts,for example,multi-scale association rules,multi-scale clustering. But it has not been involved in the field of the classification mining. This paper defined the concept of generalized fractal interpolation theory,broke the situation that limited to the use of the IFS,and extended the application of the fractal interpolation. Then,it proposed a multi-scale classification scaling-down algorithm based on the generalized fractal interpolation theory named MSCSDA. This paper performed experiments on four UCI benchmark data sets,and one real data set(H province part of the population). Then it analyzed the experimental results compare MSCSDA with KNN,decision tree and LIBSVM algorithms on different data sets. The experimental results show that the MSCSDA gives better results in terms of classification than the others.
The research of multi-scale data mining mainly applies to space remote sensing image data sets,and conducts scale division based on the resolution or regional segmentation of the images,then analysis knowledge on each scale layer. Recently,there are quite a few learners apply the multi-scale data mining to general data sets,and conduct scale division based on the level theory,concept hierarchy and inclusion degree etc.,study the distribution rule on different scale layers,and then found significant facts,for example,multi-scale association rules,multi-scale clustering. But it has not been involved in the field of the classification mining. This paper defined the concept of generalized fractal interpolation theory,broke the situation that limited to the use of the IFS,and extended the application of the fractal interpolation. Then,it proposed a multi-scale classification scaling-down algorithm based on the generalized fractal interpolation theory named MSCSDA. This paper performed experiments on four UCI benchmark data sets,and one real data set(H province part of the population). Then it analyzed the experimental results compare MSCSDA with KNN,decision tree and LIBSVM algorithms on different data sets. The experimental results show that the MSCSDA gives better results in terms of classification than the others.
引文
[1]韩玉辉,赵书良,柳萌萌,等.多尺度聚类挖掘算法[J].计算机科学,2016,43(8):244-248.(Han Yuhui,Zhao Shuliang,Liu Mengmeng,et al. Multi-scale clustering mining algorithm[J]. Computer Science,2016,43(8):244-248.)
[2]苏东海,赵书良,柳萌萌,等.基于加权向量提升的多尺度聚类挖掘算法[J].计算机科学,2015,42(4):263-267.(Su Donghai,Zhao Shuliang,Liu Mengmeng,et al. Weight vector based multi-scale clustering algorithm[J]. Computer Science,2015,42(4):263-267.)
[3]李正泉,吴尧祥.顾及方向遮蔽性的反距离权重插值法[J].测绘学报,2015,44(1):91-98.(Li Zhengquan,Wu Yaoxiang. Inverse distance weighted interpolation involving position shading[J]. Acta Geodaetica et Cartographica Sinica,2015,44(1):91-98.)
[4]梁永忠,葛咏,王江浩.基于地统计学的尺度下推方法综述[J].遥感技术与应用,2015,30(1):1-7.(Liang Yongzhong,Ge Yong,Wang Jianghao. Review of geostatistical-based downscaling[J]. Remote Sensing Technology and Application,2015,30(1):1-7.)
[5] Wang Qunming,Shi Wenzhong,Atkinson P M,et al. A new geostatistical solution to remote sensing image downscaling[J]. IEEE Trans on Geoscience and Remote Sensing,2015,54(1):1-11.
[6] Liu Liqiang,Wang Xiangguo,Ren Huili. 3D seabed terrain establishment based on moving fractal interpolation[C]//Proc of the 7th International Joint Conference on Computational Sciences and Optimization. Washington DC:IEEE Computer Society,2014:6-10.
[7]刘红艳,陈宇坤,李信富.地震道重建和重采样的分形插值方法研究[J].地球物理学进展,2014,29(2):518-522.(Liu Hongyan,Chen Yukun,Li Xinfu. Fractal interpolation method for reconstruction and resampling of seismic data[J]. Progress in Geophysics,2014,29(2):518-522.)
[8]赵泉华,刘冬,李晓丽,等.利用包含度和隶属度的遥感影像模糊分割[J].中国图象图形学报,2017,22(7):988-995.(Zhao Quanhua,Liu Dong,Li Xiaoli,et al. Fuzzy segmentation algorithm for remote sensing images using inclusion degree and membership degree[J]. Journal of Image and Graphics,2017,22(7):988-995.)
[9]李晓晖,袁峰,贾蔡,等.基于反距离加权和克里格插值的S-A多重分形滤波对比研究[J].测绘科学,2012,37(3):87-89,46.(Li Xiaohui,Yuan Feng,Jia Cai,et al. Contrastive study of S-A multifractal filtering method based on inverse distance weighted and Kriging interpolation[J]. Science of Surveying and Mapping,2012,37(3):87-89,46.)
[10]Chen Wenyen,Song Yangqiu,Bai Hongjie,et al. Parallel spectral clustering in distributed systems[J]. IEEE Trans on Pattern Analysis&Machine Intelligence,2011,33(3):568-586.
[11]Allab K,Labiod L,Nadif M. A semi-NMF-PCA unified framework for data clustering[J]. IEEE Trans on Knowledge and Data Engineering,2017,29(1):2-16.
[12] Miller D,Soh L K. Cluster-based boosting[J]. IEEE Trans on Knowledge and Data Engineering,2015,27(6):1491-1504.
[2]苏东海,赵书良,柳萌萌,等.基于加权向量提升的多尺度聚类挖掘算法[J].计算机科学,2015,42(4):263-267.(Su Donghai,Zhao Shuliang,Liu Mengmeng,et al. Weight vector based multi-scale clustering algorithm[J]. Computer Science,2015,42(4):263-267.)
[3]李正泉,吴尧祥.顾及方向遮蔽性的反距离权重插值法[J].测绘学报,2015,44(1):91-98.(Li Zhengquan,Wu Yaoxiang. Inverse distance weighted interpolation involving position shading[J]. Acta Geodaetica et Cartographica Sinica,2015,44(1):91-98.)
[4]梁永忠,葛咏,王江浩.基于地统计学的尺度下推方法综述[J].遥感技术与应用,2015,30(1):1-7.(Liang Yongzhong,Ge Yong,Wang Jianghao. Review of geostatistical-based downscaling[J]. Remote Sensing Technology and Application,2015,30(1):1-7.)
[5] Wang Qunming,Shi Wenzhong,Atkinson P M,et al. A new geostatistical solution to remote sensing image downscaling[J]. IEEE Trans on Geoscience and Remote Sensing,2015,54(1):1-11.
[6] Liu Liqiang,Wang Xiangguo,Ren Huili. 3D seabed terrain establishment based on moving fractal interpolation[C]//Proc of the 7th International Joint Conference on Computational Sciences and Optimization. Washington DC:IEEE Computer Society,2014:6-10.
[7]刘红艳,陈宇坤,李信富.地震道重建和重采样的分形插值方法研究[J].地球物理学进展,2014,29(2):518-522.(Liu Hongyan,Chen Yukun,Li Xinfu. Fractal interpolation method for reconstruction and resampling of seismic data[J]. Progress in Geophysics,2014,29(2):518-522.)
[8]赵泉华,刘冬,李晓丽,等.利用包含度和隶属度的遥感影像模糊分割[J].中国图象图形学报,2017,22(7):988-995.(Zhao Quanhua,Liu Dong,Li Xiaoli,et al. Fuzzy segmentation algorithm for remote sensing images using inclusion degree and membership degree[J]. Journal of Image and Graphics,2017,22(7):988-995.)
[9]李晓晖,袁峰,贾蔡,等.基于反距离加权和克里格插值的S-A多重分形滤波对比研究[J].测绘科学,2012,37(3):87-89,46.(Li Xiaohui,Yuan Feng,Jia Cai,et al. Contrastive study of S-A multifractal filtering method based on inverse distance weighted and Kriging interpolation[J]. Science of Surveying and Mapping,2012,37(3):87-89,46.)
[10]Chen Wenyen,Song Yangqiu,Bai Hongjie,et al. Parallel spectral clustering in distributed systems[J]. IEEE Trans on Pattern Analysis&Machine Intelligence,2011,33(3):568-586.
[11]Allab K,Labiod L,Nadif M. A semi-NMF-PCA unified framework for data clustering[J]. IEEE Trans on Knowledge and Data Engineering,2017,29(1):2-16.
[12] Miller D,Soh L K. Cluster-based boosting[J]. IEEE Trans on Knowledge and Data Engineering,2015,27(6):1491-1504.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |