文章基本信息

标题：Integrating Clustering Method in Compactly Supported Radial Basis Function for Surface Approximation
本地全文：下载
作者：Khang Jie Liew ; Kah Heng Tee ; Ahmad Ramli 等
期刊名称：IAENG International Journal of Computer Science
印刷版ISSN：1819-656X
电子版ISSN：1819-9224
出版年度：2019
卷号：46
期号：1
页码：38-45
出版社：IAENG - International Association of Engineers
摘要：Radial basis function, one of the mesh-free methods,makes it convenient to interpolate and approximate high dimensionaldata points. Interpolation fits the data points exactly,whereas approximation fits the data points approximately. Theapproximation method is more appropriate for large and noisydata points in comparison to the interpolation method. Compactlysupported radial basis function is extensively discussed inthe literature of approximation theory. It also becomes popularas a result of its computational advantages. Hence, in thisstudy, Wendland’s compactly supported radial basis function ischosen. The main contribution of this paper is integrating theclustering procedures to determine the recommended number ofreference points, which is expected to provide as few referencepoints as possible for surface approximation. Three bivariatetest functions are selected to generate a moderately largeamount of data points followed by the process of adding differentlevels of noise in order to observe the effectiveness of theproposed method. The results obtained are further confirmedand analysed through error analysis. Finally, the experimentalresults show that the surfaces are well approximated fromthe recommended number of reference points gained from theproposed method.
关键词：Compact support; Radial basis function; Clustering;Reference points; Surface approximation