The discipline of pattern recognition is traditionally divided into the statistical and the structural approach. Statistical pattern recognition is characterized by representing objects by means of feature vectors, while the structural approach uses symbolic data structures, such as strings, trees, and graphs. In this talk, we review some advances in the field of graph-based pattern recognition that aim at making algorithmic tools originally developed in statistical pattern recognition available for graphs. In particular, more recent approaches, such as graph kernels and graph embedding are discussed. These novel approaches provide very elegant and systematic ways to make the complete arsenal of statistical pattern recognition tools available for graphs. Also, we address computational complexity problems that may arise with graph representations, and how they can be possibly overcome.
Horst Bunke is King-Sun Fu Prize Award Winner 2010