Boosting support vector machines for text classification through parameter-free threshold relaxation.
James G. ShanahanNorbert RomaPublished in: CIKM (2003)
Keyphrases
- parameter free
- text classification
- feature selection
- support vector
- learning machines
- large margin classifiers
- soft margin
- naive bayes
- categorical data
- bag of words
- support vector machine
- outlier detection
- ensemble learning
- machine learning
- text categorization
- generalization ability
- labeled data
- multi class
- logistic regression
- classification accuracy
- base classifiers
- data cleaning
- fully automatic
- feature space
- svm classifier
- dimensionality reduction
- cost sensitive
- ensemble methods
- loss function
- cross validation
- text mining
- knn
- learning algorithm
- kernel function
- unsupervised learning
- feature set
- information extraction
- text classifiers
- boosting algorithms
- multi label
- data sets
- binary classification
- training data
- data mining