Finite-sample analysis of impacts of unlabeled data and their labeling mechanisms in linear discriminant analysis.
Kenichi HayashiKeiji TakaiPublished in: Commun. Stat. Simul. Comput. (2017)
Keyphrases
- unlabeled data
- linear discriminant analysis
- semi supervised learning
- active learning
- semi supervised
- labeled data
- face recognition
- principal component analysis
- training data
- co training
- text classification
- dimensionality reduction
- dimension reduction
- class labels
- high dimensional data
- unsupervised learning
- computer vision
- supervised learning
- training set
- data analysis
- class distribution
- domain adaptation
- sample size
- small sample size
- support vector machine svm
- multi class
- upper bound
- pattern recognition
- support vector
- feature extraction
- decision trees
- feature selection