Necessary and sufficient conditions for Bayes risk consistency of a recursive kernel classification rule.
Wlodzimierz GreblickiMiroslaw PawlakPublished in: IEEE Trans. Inf. Theory (1987)
Keyphrases
- sufficient conditions
- bayes risk
- classification rules
- reproducing kernel hilbert space
- loss function
- kernel methods
- genetic programming
- kernel function
- decision trees
- upper and lower bounds
- nearest neighbor
- classification algorithm
- euclidean space
- association rules
- regularization parameter
- real valued
- fuzzy rules
- distance measure
- special case
- gaussian process
- kernel matrix
- distortion measure
- learning problems
- density estimation
- data sets
- support vector machine
- input space
- linear model
- pattern recognition
- domain adaptation
- support vector
- genetic algorithm
- machine learning