A new classification strategy for human activity recognition using cost sensitive support vector machines for imbalanced data.
Bilal M'hamed AbidineBelkacem FerganiMourad OussalahLamya FerganiPublished in: Kybernetes (2014)
Keyphrases
- cost sensitive
- imbalanced data
- class imbalance
- class distribution
- human activities
- binary classification
- support vector
- support vector machine
- cost sensitive learning
- misclassification costs
- multi class
- minority class
- active learning
- svm classifier
- classification accuracy
- feature selection
- naive bayes
- decision boundary
- support vector machine svm
- multi class classification
- hyperplane
- classification models
- base classifiers
- kernel function
- decision trees
- action recognition
- generalization ability
- training set
- ensemble classifier
- sampling methods
- classification algorithm
- binary classifiers
- data sets
- benchmark datasets
- image classification
- supervised learning
- feature space
- machine learning