Table 2.

Classification algorithms and their parameter settings.

Algorithm	Parameter Setting
K-Nearest Neighbors	number of neighbors and instance weighting methods
Naïve Bayes	kernel density estimator
Bayes Net	search algorithm and estimator algorithm
Naïve Bayes Multinomial	default parameter setting in Weka
Logistic	kernel type and the corresponding parameters of each kernel type
Multilayer Perceptron	number of hidden layers, number of nodes in each layer, learning rate, and momentum
Simple Logistic	default parameter setting in Weka
Stochastic Gradient Descent	learning rate, lambda and loss function
Decision Table	attribute search method
J48	minimum number of instances per leaf, reduced error pruning and confidence threshold for pruning
Random Forest	number of trees, maximum depth of the trees, and number of attributes
Support Vector Machine	kernel type and the corresponding parameters of each kernel type