Determining the Training Window for Small Sample Size Classification with Concept Drift.
Indre ZliobaiteLudmila I. KunchevaPublished in: ICDM Workshops (2009)
Keyphrases
- small sample size
- concept drift
- classification algorithm
- labelled data
- sample size
- data stream classification
- high dimensionality
- linear discriminant analysis
- data streams
- high dimensional
- face recognition
- training set
- microarray data
- training samples
- feature selection
- drift detection
- novelty detection
- machine learning
- decision trees
- supervised learning
- drifting concepts
- non stationary
- classification accuracy
- high dimensional data
- pattern recognition
- training examples
- class labels
- sliding window
- model selection
- data distribution
- change detection
- support vector
- text classification
- dimensionality reduction
- neural network
- computer vision
- feature space
- feature vectors
- image registration
- multi dimensional
- principal component analysis
- clustering method
- support vector machine svm