max_depth=5
for setting the max depth of the Gradient Boosted Tree.
learning_rate=0.01
for setting the learning rate of the boosted tree.
min_child_weight=1
for setting minimum sum of hessian weight needed in a child node.
n_estimators=100
sets the number of Gradient Boosted trees to fit to the dataset.
Both the models were put in a Voting Classifier with a weight of 4:6 with majority in favor of XGBoost Classifier. The accuracy for the model was checked for 5-fold cross-validation and the accuracy comes to around 77.63% and an ROC score of 0.78.