文章基本信息

标题：Self-organization of Polynomial Regression Models in Neural Structures of Geometric Transformations
本地全文：下载
作者：Roman Tkachenko ; Sergii Demianchuk
期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2016
卷号：5
期号：4
页码：0919-0923
出版社：Shri Pannalal Research Institute of Technolgy
摘要：In this paper presented methods for building self-organized polynomial regression models using input signal's functional expansion. The aim is to produces better prediction results using new developed prediction methods, compared to the known algorithms, prediction based on auto-associative method, the KNN method (k-nearest neighbors), and the group method of data handling (GMDH). Input signal's functional expansion is performed using the set of Kolmogorov-Gabor polynomials. Principal components are used to construct polynomial of the Kolmogorov-Gabor. The results of using newly developed methods prove effectiveness of the forecasting ability for the samples of a large size.
关键词：polynomial regression models; neural ; network; auto associative neural network; group method of ; data handling; model of geometric transformation