A Dynamic Early Stopping Criterion for Random Search in SVM Hyperparameter Optimization.
Adrian-Catalin FloreaRazvan AndoniePublished in: AIAI (2018)
Keyphrases
- random search
- grid search
- early stopping
- support vector machine
- parameter optimization
- cross validation
- hyperparameters
- simulated annealing
- search space
- stochastic gradient descent
- support vector
- random forest
- genetic algorithm
- model selection
- support vector machine svm
- gaussian process
- machine learning
- incremental learning
- optimization algorithm
- genetic algorithm ga
- feature selection
- active learning
- incomplete data
- training data
- feature space
- learning algorithm
- knn
- random sampling
- multi class
- parameter settings
- matrix factorization
- convergence rate