Variational Bayes for continuous hidden Markov models and its application to active learning

Mathematics

• Hidden Markov Models 100%
• Minimizes 66%
• Posteriori 33%
• Training Procedure 33%
• Learned Model 33%
• Variance 33%
• Maximum Likelihood 33%
• Information Gain 33%
• Test Data 33%
• Training Set 33%
• Point Estimate 33%
• Feature Vector 33%

Computer Science

• Hidden Markov Models 100%
• Active Learning 100%
• maximum-likelihood 20%
• Experimental Result 20%
• Information Gain 20%
• Learning Algorithm 20%
• Feature Vector 20%
• Classification Error 20%
• Random Selection 20%
• Point Estimate 20%
• Parameter Uncertainty 20%

Keyphrases

• Query by Committee 33%