HMM-ModE - Improved classification using profile hidden Markov models by optimising the discrimination threshold and modifying emission probabilities with negative training sequences.
Prashant K. SrivastavaDhwani K. DesaiSoumyadeep NandiAndrew M. LynnPublished in: BMC Bioinform. (2007)
Keyphrases
- profile hidden markov models
- hidden markov models
- sequence classification
- sequence data
- discriminative training
- training samples
- supervised learning
- training set
- classification accuracy
- feature selection
- pattern recognition
- training process
- training phase
- multi layer perceptron
- discriminative classifiers
- feature vectors
- classification scheme
- classification models
- pattern classification
- text classification
- decision boundary
- support vector
- sequential data
- machine learning
- classification algorithm
- support vector machine svm
- support vector machine
- decision trees
- decision rules
- roc curve
- speech recognition
- training dataset
- class labels
- training examples
- image classification
- probability estimates
- active learning
- training data
- estimate the model parameters
- data sets