Sets of best performing parameters obtained after gridsearching through different values for each classification algorithm.

Selection was based on the highest 10-fold cross-validation score. Parameters tuned for SVM were Cost function CSVM, Kernel and Gamma value. Parameters considered for RF were No. estimators, Max depth, Min sample leaf and Min sample split. For LogR the parameters considered were Cost Function CLR and solver. For KNN, No. of neighbours and Metric Distance was considered for tuning. The CV score shows the mean accuracy score of the best classifier obtained with the corresponding parameters.