Linear Models for Structure Prediction

Over the last few years, several groups have been developing models and algorithms for learning to predict the structure of complex data, sequences in particular, that extend well-known linear classification models and algorithms, such as logistic regression, the perceptron algorithm, and support vector machines. These methods combine the advantages of discriminative learning with those of probabilistic generative models like HMMs and probabilistic context-free grammars. I introduce linear models for structure prediction and their simplest learning algorithms, and exemplify their benefits with applications to information extraction from biomedical text, dependency parsing of English and Czech, and gene finding.

Joint work with Koby Crammer, Ryan McDonald, Fei Sha (University of Pennsylvania), Hanna Wallach (Cambridge University), in collaboration with Andrew McCallum and his group at the University of Massachusetts and John Lafferty at CMU, funded by NSF (EIA 0205448, EIA 0205456, IIS 0428193) and DARPA (SRI contract NBCHD030010).