Finding optimal neural networks for land use classification

In this communications, we present a fully automatic and computationally efficient algorithm based on the minimum description length principle (MDL) for optimizing multilayer perceptron (MLP) classifiers. We demonstrate our method on the problem of multispectral Landsat image classification. We compare our results with a hand-designed MLP and a Gaussian maximum likelihood classifier, in which our method produces better classification accuracy with a smaller number of hidden units.