Credal model averaging for classification: representing prior ignorance and expert opinions.
Giorgio CoraniAndrea MignattiPublished in: Int. J. Approx. Reason. (2015)
Keyphrases
- model averaging
- bayesian methods
- supervised classification
- expert opinions
- decision trees
- prior knowledge
- classification accuracy
- posterior distribution
- supervised learning
- text classification
- machine learning
- training set
- feature vectors
- support vector
- machine learning methods
- document classification
- model selection
- sample size
- least squares
- maximum a posteriori
- gaussian processes
- class labels
- learning algorithm