Structural Perceptrons for Attributed Graphs
We propose a structural perceptron for supervised and unsupervised learning on data represented in terms of attributed graphs. We integrate structural perceptrons into a multi-layer perceptron and competitive learning network to provide examples of supervised and unsupervised neural learning machines which are suited to process graphs. In first experiments the proposed algorithms were successfully applied to function regression, classification, and clustering.
- 3.Geibel, P., Jain, B., Wysotzki, F.: SVM learning with the SH inner product. In: Verleysen, M. (ed.) Proc. of the 12th European Symposium on Artificial Neural Networks, ESANN 2004, D-Facto, Brussels (2004) (accepted for publication)Google Scholar
- 5.Herbrich, R.: Learning Kernel Classifiers. The MIT Press, Cambridge (2002)Google Scholar
- 7.Jain, B., Wysotzki, F.: Central clustering in the domain of graphs. Machine Learning Journal (2004) (accepted for publication)Google Scholar