Optimizing Training Data and Hyperparameters of Support Vector Machines Using a Memetic Algorithm.
Wojciech DudzikMichal KawulokJakub NalepaPublished in: ICMMI (2019)
Keyphrases
- hyperparameters
- memetic algorithm
- support vector
- training data
- cross validation
- model selection
- classification accuracy
- tabu search
- training set
- bayesian framework
- evolutionary computation
- incomplete data
- bayesian inference
- closed form
- support vector machine
- em algorithm
- learning algorithm
- decision trees
- data sets
- kernel function
- random sampling
- decision function
- noise level
- sample size
- prior knowledge
- combinatorial optimization
- vehicle routing problem
- incremental learning
- genetic algorithm
- maximum likelihood
- prior information
- supervised learning
- crossover operator
- genetic programming
- expectation maximization
- feature selection
- missing values
- maximum a posteriori
- support vector regression
- evolutionary algorithm
- bayesian networks