Combining a multi-objective optimization approach with meta-learning for SVM parameter selection.
Péricles B. C. de MirandaRicardo Bastos Cavalcante PrudêncioAndré Carlos Ponce de Leon Ferreira de CarvalhoCarlos SoaresPublished in: SMC (2012)
Keyphrases
- multi objective optimization
- parameter selection
- meta learning
- model selection
- multi objective
- evolutionary algorithm
- feature selection
- inductive learning
- learning tasks
- ls svm
- support vector machine svm
- support vector
- genetic algorithm
- cross validation
- machine learning algorithms
- machine learning
- support vector machine
- differential evolution
- metamodel
- data mining
- fitness function
- optimization algorithm
- feature vectors
- data sets
- decision trees
- hyperparameters
- base classifiers
- knn
- multi class
- kernel methods
- markov random field
- feature space
- variable selection
- svm classifier
- least squares
- k nearest neighbor
- training set
- generalization ability
- bayesian networks
- particle swarm optimization
- kernel function