Random forest is an ensemble learning method based on the construction of a multitude of decision trees. Each individual tree in the random forest spits out a class prediction and the class with the most votes (mode) becomes our model’s prediction. This specific random forest algorithm is designed for the binary classification tasks.
To use the algorithm you need to specify the learning rate, the number of iterations, the number of leaves.
- Note that higher number of leaves leads to increased accuracy, but rises the chances of overfitting
- The number of the boosting iterations is suggested to be set inversely to the learning rate (decrease one while increasing second).