Notes. The search parameter space indicates the values in between which the optimal value is sought, and the optimal value is displayed on the right. To find the optimal value, we used a Bayesian optimisation algorithm. num_leaves is the maximum number of leaves a tree can have. learning_rate determines the step size during the learning process. min_data_in_leaf is the minimum number of samples in each decision leaf. colsample_bytree is the random subset of features the model trains on each iteration. reg_alpha and reg_lambda are L1 and L2 regularisation, respectively.