Selecting parameters of SVM using meta-learning and kernel matrix-based meta-features.
Carlos SoaresPavel BrazdilPublished in: SAC (2006)
Keyphrases
- meta learning
- kernel matrix
- feature space
- kernel function
- kernel methods
- feature vectors
- support vectors
- learning tasks
- svm classifier
- feature selection
- support vector
- support vector machine
- model selection
- support vector machine svm
- inductive learning
- low rank
- metric learning
- multiple kernel learning
- feature set
- image features
- input space
- ls svm
- training examples
- training samples
- principal component analysis
- collaborative filtering
- matrix factorization
- least squares
- nearest neighbor
- hyperparameters
- knn
- classification accuracy
- machine learning
- data mining