A PAC-Bayesian approach to minimum perplexity language modeling

Keyphrases

• Bayesian Approach 100%
• Language Modeling 100%
• Sequence Clustering 100%
• PAC-Bayesian 100%
• Unsupervised Learning 50%
• Speech Recognition 50%
• High Probability 50%
• Clustering Algorithm 50%
• Sparsification 50%
• Generalization Error 50%
• Resource Management 50%
• Language Model 50%
• Machine Translation 50%
• Class-based 50%
• Statistical Language Model 50%
• N-gram 50%
• HMM-based 50%
• N-gram Model 50%
• Cluster Assignment 50%
• Assignment Probabilities 50%
• Vocabulary Size 50%
• Popular Language 50%
• Bayesian Theorem 50%
• Probability Guarantee 50%

Computer Science

• Bayesian Approach 100%
• Language Modeling 100%
• Unsupervised Learning 33%
• Speech Recognition 33%
• Clustering Algorithm 33%
• Machine Translation 33%
• n-gram language model 33%
• Generalization Error 33%