This lecture focuses on the problem of training structured classifiers. For the purpose of this lecture, we will assume that we have access to an inference algorithm that can perform combinatorial optimization for us. We will primarily look at structured versions of the Perceptron and the SVM algorithms under the empirical risk minimization framework.
Lectures
Readings
-
Chapter 3 of Noah Smith, Linguistic Structure Prediction. (required)
-
Nathan Srebro and Ambuj Tewari, Stochastic Optimization for Machine Learning, Slides of tutorial presented at ICML 2010. (required)
-
Keerthi, S. Sathiya, and Sellamanickam Sundararajan. CRF versus SVM-struct for sequence labeling. Technical report, Yahoo Research, 2007.
-
(*) Fei Sha and Fernando Pereira, Shallow Parsing with Conditional Random Fields, NAACL 2003.
-
(*) Ioannis Tsochantaridis, Thomas Hofmann, Thorsten Joachims, and Yasemin Altun. Support vector machine learning for interdependent and structured output spaces, ICML 2004.
-
(*) Ben Taskar, Carlos Guestrin, and Daphne Koller. Max-margin Markov networks, NIPS 2004.
-
(*) Vasin Punyakanok, Dan Roth and Wen-tau Yih, The Importance of Syntactic Parsing and Inference in Semantic Role Labeling, Computational Linguistics 2008