Employing unlabeled data to improve the classification performance of SVM, and its application in audio event classification.
Yan LengChengli SunXinyan XuQi YuanShuning XingHonglin WanJingjing WangDengwang LiPublished in: Knowl. Based Syst. (2016)
Keyphrases
- improve the classification accuracy
- unlabeled data
- supervised learning
- support vector
- support vector machine svm
- class labels
- support vector machine
- semi supervised classification
- classification algorithm
- text classification
- feature selection
- labeled data
- training set
- semi supervised
- classification systems
- feature space
- unsupervised learning
- training data
- active learning
- classification accuracy
- pattern recognition
- feature extraction
- semi supervised learning
- supervised learning algorithms
- machine learning
- svm classifier
- decision boundary
- training examples
- image classification
- reinforcement learning
- text categorization
- feature set
- pattern classification
- cost sensitive
- knn
- labeled and unlabeled data
- decision trees
- information retrieval
- data sets
- partially supervised