A Novel Method for Partitioning Feature Spaces According to their Inherent Classification Complexity.
Giuliano ArmanoFrancesco MasciaPublished in: Int. J. Pattern Recognit. Artif. Intell. (2013)
Keyphrases
- support vector machine svm
- classification accuracy
- feature space
- classification method
- training samples
- support vector machine
- pattern classification
- preprocessing
- classification algorithm
- feature set
- computational cost
- classification process
- cost function
- significant improvement
- cross validation
- benchmark data sets
- high accuracy
- feature extraction
- input data
- classification scheme
- dimensionality reduction
- machine learning methods
- detection method
- clustering method
- unsupervised learning
- model selection
- dynamic programming
- image retrieval
- image segmentation
- neural network
- probabilistic model
- decision rules
- high dimensional
- support vector
- similarity measure
- training phase
- machine learning