文章基本信息

标题：Classification of Spatial Data Using a Hybrid Neural Network-Expert System
本地全文：下载
作者：Andrew K. Skidmore
期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：1992
卷号：XXIX Part B7
页码：1015-1018
出版社：Copernicus Publications
摘要：A hybrid neural network/expert system has been developed to~a'P land cc:>ver such as forest soil or vegetation overstorey fromdlg11:~ spatial data (that ~ay include remotely sensed imagery,elevatIOn data, and terram features derived from the elevationdata). The method is different because the hybrid systemexhibits the inherent advantages of both neural network andexpert system models. Firstly, the hybrid system may acquire(or lea.rJ?) geograp~ic and forest management knowledge from~he spat~al data, usmg homo~eneous training areas. Secondly, itIS possIble to record and mterrogate "experience" using aknowledge base linked to the neural network, and to use theknowledge to assist in the classification of land cover. Theexpert system and neural network components of the hybridsystem operate as a fully integrated system; in other words theexpert system does not simply process the output of the n~uralnetwork.
关键词：Artificial Intelligence; Expert Systems; Neural;Networks; Renewable Resources; Classification