Combining multiple classifiers: An application using spatial and remotely sensed information for land cover type mapping

This article discusses two new methods for increasing the accuracy of class ifiers used land cover mapping. The first method, called the product rule, is a simple and general method of combining two or more classification rule s as a single rule. Stacked regression methods of combining classification rules are discussed and compared to the product rule. The second method of increasing classifier accuracy is a simple nonparametric classifier that us es spatial information for classification. Two data sets used for land cove r mapping of Landsat TM scenes from Idaho and Montana illustrate the new me thods. For these examples, the product rule compared favorably to the more complex stacked regression methods. The spatial classifier produced substan tial increases in estimated accuracy when combined with one or more additio nal classifiers that used remotely sensed variables for classification. The se results suggest that the product rule may produce increases in map accur acy with little additional expense or effort. The spatial classifier may be useful for increasing accuracy when patterns exist in the spatial distribu tion of land cover. (C) Elsevier Science Inc., 2000.