Evaluating hyper-parameter tuning using random search in support vector machines for software effort estimation.
Leonardo Villalobos-AriasChristian Quesada-LópezJose Guevara-CotoAlexandra MartínezMarcelo JenkinsPublished in: PROMISE (2020)
Keyphrases
- parameter tuning
- random search
- support vector
- learning machines
- large margin classifiers
- simulated annealing
- parameter optimization
- search space
- ink bleed
- genetic algorithm
- parameter settings
- kernel function
- cross validation
- learning algorithm
- model selection
- generalization ability
- hyperparameters
- support vector machine
- evolutionary algorithm
- search algorithm
- machine learning
- least squares
- fuzzy logic
- kernel parameters