From: An adaptive hybrid african vultures-aquila optimizer with Xgb-Tree algorithm for fake news detection
Classification algorithm | Parameters |
---|---|
Xgb-Tree | Number of boosting iterations \(nrounds = 100\) |
Maximum depth of a tree \(max\_depth = 3\) | |
Minimum loss reduction \(gamma = 0\) | |
Minimum sum of instance weight \(min\_child\_weight = 1\) | |
Step size shrinkage (learning rate) \(eta = 0.4\) | |
Sub-sample ratio of columns \(colsample\_bytree = 0.8\) | |
Sub-sample ratio of training \(sub\_sample = 0.75\) | |
DT | Maximum depth of a tree \(max\_depth = 5\) |
Number of features \(max\_features = 1\) | |
k-NN | Euclidean distance metric \(k=5\) |
SVM | Regularization parameter \(C=1\) |
Degree of polynomial kernel \(degree=2\) | |
RF | Number of trees in a forest \(n\_estimators=10\) |
Maximum depth of a tree \(max\_depth = 5\) | |
Number of features \(max\_features = 1\) | |
MLP | Number of neurons in the \(i^{th}\) hidden layer \(hidden\_layer\_sizes = (1000,500,100)\) |
Strength of the L2 regularization term \(alpha = 0.001\) | |
Maximum number of iterations \(max\_iter = 1000\) |