Support vector machines - Part 2: Practical examples and applications [Support vector machines teil 2: Praktische beispiele und anwendungen]

摘要

The support vector machines (SVM) offer a lot of possibilities to model the results of geodetic observations. The theoretical overview of SVM including their derivation has already been presented within the first part of this publication. Herein, the linear and nonlinear algorithms for pattern recognition and regression purposes have been explained. This available second part is focussed on the practice of SVM computations. Accordingly, small computational examples will show, on the one hand, how these algorithms work. On the other hand, the adaptation of SVM on the monitoring of recently active tectonics or of dangerous landslides will demonstrate their scope of use.