A binary-encoded tabu-list genetic algorithm for fast support vector regression hyper-parameters tuning.
J. Gascón-MorenoSancho Salcedo-SanzEmilio G. Ortíz-GarcíaLeopoldo Carro-CalvoBeatriz Saavedra-MorenoJosé Antonio Portilla-FiguerasPublished in: ISDA (2011)
Keyphrases
- support vector regression
- hyperparameters
- support vector
- genetic algorithm
- tabu list
- tabu search
- parameter optimization
- model selection
- cross validation
- grid search
- regression model
- closed form
- simulated annealing
- search strategy
- bayesian framework
- hybrid algorithm
- prior information
- parameter settings
- bayesian inference
- random sampling
- metaheuristic
- kernel function
- sample size
- support vector machine
- job shop scheduling problem
- maximum likelihood
- incremental learning
- maximum a posteriori
- em algorithm
- evolutionary algorithm
- multi objective
- noise level
- neural network
- fitness function
- support vector machine svm
- scheduling problem
- search algorithm
- ls svm
- genetic algorithm ga
- artificial neural networks
- image processing
- classification accuracy
- incomplete data
- feature vectors
- similarity measure
- pattern recognition
- computer vision
- pairwise
- feature selection
- prior knowledge
- active learning
- semi supervised
- genetic programming
- differential evolution