Fig 9. Number of decision trees and OOB error rate curve.

Fig 9 shows that the classification error rate decreases with an increase in the number of decision trees. After reaching the 100 trees, the classification error trends to be stable (approximately 0.01).