The support vector machine (SVM) classification algorithm often achieves quite high accuracy on hyperspectral images, even when trained on small amounts of data. However, SVMs can still be computationally expensive relative ...