文章基本信息

标题：Dimensional Reduction of Statistical Structural of a Paper by Information Geometry
本地全文：下载
作者：Dr K Ramesh Babu ; K Srinivasa Rao ; P.Prasanna Kumar 等
期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：1
页码：746
DOI：10.15680/IJIRCCE.2017.0501159
出版社：S&S Publications
摘要：the visual and especially tactile quality of a surface is called a texture. Stochastic textures with featuresspanning many length scales arise in a range of contexts in sciences, from nano size structures like synthetic bone to aocean wave height distributions and cosmic phenomena like inter-galactic cluster void distributions. The samples herecame from the papermaking industry but such a reduction of large frequently noisy spatial data sets is useful in a rangeof materials and contexts at all scales. Aim is reducing the size of structure without any loss in information by usinginformation geometry. We are giving answer to “how far apart are two distributions?” e.g. Gaussian (m, s2 ): Euclidiandistance between two distributions has no `natural’ statistical significance. Information geometry seeks first the shapeof the (multidimensional) surface and once the surface is known the shortest curve between two points representing thedistributions is the ‘natural’ metric, Information geometry provides a natural metric to discriminate among formationtextures.
关键词：texture; Dimensionality reduction; information metric; statistical etc