Novel multi-class feature selection methods using sensitivity analysis of posterior probabilities.
Kai Quan ShenChong Jin OngXiao Ping LiEinar P. V. Wilder-SmithPublished in: SMC (2008)
Keyphrases
- sensitivity analysis
- multi class
- posterior probability
- bayesian networks
- bayesian framework
- generative model
- managerial insights
- probability distribution
- support vector machine
- multiple classes
- multi class classification
- cost sensitive
- conditional probabilities
- pairwise
- probability density function
- bayes theorem
- probabilistic model
- feature selection
- binary classification
- error correcting output codes
- binary and multi class
- binary classifiers
- prior knowledge
- multi class boosting
- learning algorithm
- machine learning