Mejora del desempeño de SVM usando un GA y la creación de puntos artificiales para conjuntos de datos no balanceados.
José Hernández SantiagoJair Cervantes CanalesCarlos Hiram Moreno MontielBeatriz Hernández SantiagoPublished in: Res. Comput. Sci. (2016)
Keyphrases
- genetic algorithm ga
- support vector machine svm
- support vector
- genetic algorithm
- support vector machine
- knn
- multi class
- multi class classification
- svm classification
- kernel function
- optimal parameters
- fitness function
- hybrid algorithm
- generalization ability
- training data
- feature selection
- hybrid particle swarm optimization
- hyperplane
- support vectors
- machine learning
- svm classifier
- neural network
- artificial neural networks
- classification algorithm
- multi class svm
- parallel genetic algorithm
- decision function
- text classifiers
- ant colony optimization
- kernel methods
- classification method
- multi objective
- training process
- particle swarm optimization
- feature space
- penalty function
- similarity measure
- k nearest neighbor
- parameter set
- hybrid ga
- messy genetic algorithm