Kernel-based discriminative learning algorithms for labeling sequences, trees, and graphs

Abstract

We introduce a new perceptron-based discriminative learning algorithm for labeling structured data such as sequences, trees, and graphs. Since it is fully kernelized and uses pointwise label prediction, large features, including arbitrary number of hidden variables, can be incorporated with polynomial time complexity. This is in contrast to existing… (More)
DOI: 10.1145/1015330.1015383

9 Figures and Tables

Topics

  • Presentations referencing similar topics