Data-driven multinomial random forest: a new random forest variant with strong consistency

Journal of Big Data

Table 4 Mean square error(%) of different RFs on benchmark datasets

Datasets	DMRF	MRF(SE)	MRF(b)	BRF(SE)	BRF(b)	Denil14(SE)	Denil14(b)	BreimanRF
ALE	0.0474	0.0758	0.0494	0.0921	0.4890	0.0931	0.0516	0.0423*
Alcohol	0.0023*	0.0068	0.0025	0.0049	0.0025	0.0033	0.0024	0.0028
Servo	0.3644	0.4248	0.2543	0.6864	0.4247	0.7832	0.2538	0.3114*
CSM	0.0204*	0.0711	0.0206	0.1681	0.0209	0.0495	0.0244	0.0239
Real estate	55.060	106.919	55.628	118.417	61.783	110.264	93.836	53.699*
Facebook	0.0510	0.7594	0.0637	1.0288	0.0546	0.8786	0.0546	0.0465*
Las Vegas Strip	0.9942	1.001	1.016	1.015	1.014	1.167	1.072	0.9826*
Forest fire	2.073*	2.159	2.191	2.166	2.164	2.166	2.156	2.232
ISTANBUL STOCK	2.14E-04*	3.88E-04	2.21E-04	4.15E-04	4.14E-04	4.40E-04	3.33E-04	2.20E-04
Qsar fish toxicity	0.7794*	1.047	0.7592	2.0185	0.9218	1.9508	0.9441	1.241
Concrete	40.080	128.167	27.749	256.478	36.889	259.715	42.966	23.869*
Qsar BCF Kow	0.7401	1.5096	0.7695	1.4094	0.7514	1.4111	1.1961	0.5152*
Flare	0.5219*	0.5222	0.5937	0.5426	0.5227	0.5684	0.6013	0.5981
Communities	0.0203	0.0372	0.0205	0.0521	0.2049	0.0499	0.0244	0.0182*
Skillcraft	4.37E-08	6.54E-08	4.39E-08	6.95E-08	5.34E-08	7.02E-08	5.64E-08	4.23E-08*
Winequality (white)	0.3562*	0.6177	0.3601	0.7749	0.3815	0.7311	0.3949	0.3640
Parkinsons	0.0013*	0.0029	0.0014	0.0078	0.0016	0.0045	0.0016	0.0013*
SeoulBikeData	0.2059	0.6722	0.207	1.4543	0.3664	1.2996	0.8322	0.1782*
Insurance	0.0585	0.0559	0.0596	0.0561	0.0549	0.0561	0.0601	0.0554*
Combined	11.700	15.371	11.716	48.995	12.445	26.054	14.445	10.819*
Cbm	1.00E-06	8.00E-06	1.00E-06	5.55E-05	1.50E-05	5.61E-05	2.00E-06	7.95E-07*