A New Multi-Class WSVM Classification to Imbalanced Human Activity Dataset.
M'hamed B. AbidineBelkacem FerganiPublished in: J. Comput. (2014)
Keyphrases
- multi class
- human activities
- binary classification problems
- single class
- support vector machine
- multiclass classification
- binary classifiers
- multi class classifier
- multi class classification
- protein classification
- binary and multi class
- multi class problems
- multiple classes
- multi class boosting
- cost sensitive
- multi class svm
- error correcting output codes
- human actions
- feature selection
- generalization ability
- multi class svms
- class imbalance
- binary classification
- activity recognition
- pairwise
- machine learning
- benchmark datasets
- action recognition
- protein sequences
- support vector
- object detection
- text classification
- class distribution
- classification accuracy
- binary classification tasks
- image classification
- kernel logistic regression
- feature set
- probabilistic boosting tree
- feature vectors
- feature extraction
- training data
- cost sensitive learning
- data streams
- string kernels
- knn
- multi label classification
- misclassification costs
- decision boundary
- high dimensionality
- svm classifier