A PAC-Bayesian Approach to Minimum Perplexity Language Modeling

6078

Knowledge Triples

Source Entity	Relationship	Target Entity
Class-Based N-Grams Method	evaluatedBy	Test Set Perplexity Metric
HMM Based Sequence Clustering Method	implements	Sequence Clustering Task
N-Gram Model Method	evaluatedBy	Test Set Perplexity Metric
PAC Bayesian Theorems Method	appliedIn	Unsupervised Learning Method
Sequence Clustering Task	dependsOn	Vocabulary Material
Statistical Language Models Method	usedIn	Machine Translation Task
Statistical Language Models Method	usedIn	Speech Recognition Task