A New Random Forest Algorithm Based on Learning Automata

Table 6

Friedman test statistical verification results for ranking the parameters of reward and penalty and comparing the proposed method with the literature.

Method	Tuning	Mean rank	Final rank
LRP	a = 0.5, b = 0.5	19.17	1
LRP	a = 0.3, b = 0.3	16.83	2
LRP	a = 0.7, b = 0.7	15.58	3
MV	Majority voting	14.67	4
LRP	a = 0.1, b = 0.1	13.92	5
LReP	a = 0.05, b = 0.01	12.17	6
LReP	a = 0.1, b = 0.01	11.83	7
LReP	a = 0.5, b = 0.01	10.08	8
LRP	a = 0.05, b = 0.05	9.58	9
RF	Random forest	9.17	10
LRP	a = 0.01, b = 0.01	8.75	11
LIR	a = 0.01, b = 0	8.42	12
LIR	a = 0.05, b = 0	7.67	13
LIR	a = 0.1, b = 0	7.67	13
LIR	a = 0.3, b = 0	7.67	13
LIR	a = 0.5, b = 0	7.67	13
LIR	a = 0.7, b = 0	7.67	13
AV	Averaging	7.58	14
LReP	a = 0.3, b = 0.01	7.17	15
LReP	a = 0.7, b = 0.01	6.75	16