Support Vector Machines (SVM) have gained increasing attention due to their classification accuracy, robustness and indifference towards the input data type. Thus, they are widely used in the remote sensing community — and especially among researchers working on hyperspectral datasets. However, since their first publication a lot of enhancements and adaptations have been proposed, many of which aim at introducing probability distributions and the Bayes theorem to SVM. Within this paper, we present a classification result of a HyMap dataset using two of the proposed enhancements — Import Vector Machines and Relevance Vector Machines — and compare them to the Support Vector Machine.