Classication Trees and Mixed Pixel Training Data
Chandra P. Giri · 2012
Research in the last decade on supervised land-cover classication has emphasized new distribution-free algorithms as high-performance alternatives to traditional classiers. Such classiers include decision trees, neural networks, nearest neighbor, and support vector machine algorithms. Distribution-free algorithms work on the spectral frontiers between land-cover classes, a marked improvement over conventional parametric classiers reliant on the statistics of central tendency. A number of comparisons between distribution-free methods have been made, which have historically favored parametric techniques. Ince (1987) and Hardin and Thomson (1992) showed that nearest-neighbor classiers were superior to parametric classiers. Hansen et al. (1996) and Friedl and Brodley (1997) found comparable performance between a classication tree approach and a maximum likelihood one. Key et al. (1989), Bischof et al. (1992), and Gopal et al. (1999) tested the maximum likelihood classier versus neural network classiers and found that the neural network classiers provide accuracies similar to or superior than that provided by the maximum likelihood classier. Likewise, support vector machines have been compared to the maximum likelihood classier and have been found to yield higher accuracies (Huang et al., 2002). Support vector machines, in turn, have been found to outperform decision trees and neural nets (Huang et al., 2002). However, variables such as the number of features, model parameter selection, and the number of training samples can affect the relative performance of distribution-free classiers (Pal and Mather, 2003).