A Meta-Learning Approach to Conditional Random Fields Using Error-Correcting Output Codes.
Francesco CiompiOriol PujolPetia RadevaPublished in: ICPR (2010)
Keyphrases
- meta learning
- conditional random fields
- error correcting output codes
- base classifiers
- multi class
- pairwise
- graphical models
- decision trees
- ensemble methods
- learning tasks
- inductive learning
- learning problems
- machine learning algorithms
- hidden markov models
- probabilistic model
- higher order
- model selection
- markov random field
- training set
- naive bayes
- information extraction
- feature selection
- generative model
- class labels
- binary classifiers
- machine learning
- cost sensitive
- test data
- data mining
- multiclass problems
- reinforcement learning
- training samples
- feature set
- support vector machine
- benchmark datasets
- computer vision
- learning algorithm
- binary classification
- data sets
- multi class problems
- graph cuts