Spatial neighborhood classifiers

摘要

Spatial data classification is a high-frequency spatial decision evaluation method. It can only choose according to experience frequently. When there were many spatial decision evaluation conditional attributes, such as hyperspectral image, it has obviously been lack of strong mathematics foundation. As an uncertainty mathematical method, Pawlak rough set can only dispose discrete data formerly, so we must discrete spatial continuous data when using this method, it would bring the profits and losses of the information in the transformation process. We used neighborhood rough set concept, and put forward a spatial continuous data classification method based on neighborhood rough set, when conditional attribute is continuous data and decision attribute is discrete data.