Investigating the optimization of hyperparameters using an RBF SVM, on different datasets

This paper looks at the tuning of hyperparameters for the Support Vector Machine (SVM) classifier. Methods such as Cross-Validation for the tuning of C and γ, Nested CV for an unbiased score of the classifier and ROC Curves for the analysis of the classifier will be used. This will be conducted on 5 different UCI Datasets (Dheeru & Karra Taniskidou, 2017) to see how different variables such as the number of examples and features or class distributions within datasets effect the classifiers optimal hyperparameters