Meta-learning Recommendation of Default Hyper-parameter Values for SVMs in Classification Tasks.
Rafael Gomes MantovaniAndré L. D. RossiJoaquin VanschorenAndré C. P. L. F. de CarvalhoPublished in: MetaSel@PKDD/ECML (2015)
Keyphrases
- parameter values
- meta learning
- feature selection
- support vector
- learning tasks
- model selection
- inductive learning
- machine learning
- parameter estimation
- collaborative filtering
- recommender systems
- machine learning algorithms
- support vector machine svm
- kernel function
- data mining
- parameter space
- optimal parameter values
- knn
- feature space
- support vector machine
- decision trees
- base classifiers
- generalization ability
- higher order
- multi class
- grid search
- data sets
- knowledge acquisition
- markov random field
- probabilistic model
- image processing
- learning algorithm