Tsochantaridis et al, JMLR 2005

Citation

Large Margin Methods for Structured and Interdependent Output Variables, by I Tsochantaridis, T. Joachims, T. Hofmann, Y. Altun. In Journal of Machine Learning Research, 2005.

This Paper can be found online [1].

Summary

This paper deals with the problem of learning a mapping from input vectors $x\in {\mathcal {X}}$ to discrete variables $y\in {\mathcal {Y}}$. However, ${\mathcal {Y}}$ is not a finite set of labels (as in multiclass classification) or a real number (as in regression). Instead, ${\mathcal {Y}}$ is a set of structured objects (such as trees, sequences, graphs, strings, etc).