Multivariable Regression and Adaptive Neurofuzzy Inference System Predictions of Ash Fusion Temperatures Using Ash Chemical Composition of US Coals

Table 4

Precision of the prediction of AFTs using different inputs in the regression method.

		Parameter	Input 1	Input 2	Input 3	Input 4	Input 5	Input 6	Input 7	Input 8
IDT	L	𝑅^2	0.820	0.827	0.835	0.837	0.856	0.866	0.868	0.87
		S.D.	119.39	117.65	114.78	114.27	107.421	103.49	102.77	101.93
		Mean	3.25𝐸−13	5.02𝐸−13	−1.01𝐸−12	−7.10𝐸−13	5.27𝐸−13	−1.54𝐸−13	−1.28𝐸−12	−0.9647
		Re (%) < 120°F	65.38	70.19	68.94	72.40	75.38	78.46	77.98	76.82
	NL	𝑅^2	0.887	0.839	0.883	0.88	0.865	0.889	0.89	0.891
		S.D.	95.395	113.606	96.264	97.997	103.911	94.256	93.836	93.272
		Mean	−3.61𝐸−6	1.97𝐸−8	8.09𝐸−7	−2.32𝐸−9	−3.31𝐸−8	−1.14𝐸−7	−2.19𝐸−8	−0.11
		Re (%) < 120°F	81.73	72.30	80.86	80.28	77.59	81.44	82.11	83.26
ST	L	𝑅^2	0.873	0.847	0.882	0.873	0.884	0.893	0.895	0.907
		S.D.	94.75	103.91	91.35	94.718	90.50	86.94	86.093	91.32
		Mean	−8.40𝐸−13	2.44𝐸−13	−1.20𝐸−12	−2.17𝐸−13	−1.04𝐸−12	1.68𝐸−13	−1.27𝐸−12	−0.0742
		Re (%) < 120°F	81.05	75.56	82.69	80	81.82	85.19	84.90	87.69
	NL	𝑅^2	0.914	0.862	0.909	0.903	0.874	0.916	0.917	0.917
		S.D.	78.109	98.792	79.87	82.706	100.233	77.138	76.744	76.792
		Mean	4.59𝐸−11	−2.78𝐸−11	6.36𝐸−7	−3.08𝐸−8	−6.33𝐸−9	1.37𝐸−7	3.64𝐸−8	−0.08
		Re (%) < 120°F	88.84	77.21	88.75	85.38	78.17	88.46	89.32	89.51
FT	L	𝑅^2	0.932	0.884	0.931	0.911	0.918	0.918	0.92	0.937
		S.D.	59.6	83.89	60.09	68.520	65.65	65.55	64.69	57.445
		Mean	1.99𝐸−13	6.02𝐸−13	−7.01𝐸−13	−3.33𝐸−13	1.96𝐸−13	−3.67𝐸−13	−6.68𝐸−13	−0.0532
		Re (%) < 120°F	98.75	84.31	97.59	94.13	95.19	95.09	95.19	98.07
	NL	𝑅^2	0.94	0.889	0.934	0.922	0.876	0.934	0.936	0.94
		S.D.	56.348	79.086	58.95	64	95.549	58.944	57.911	56.118
		Mean	2.28𝐸−10	2.27𝐸−10	−1.17𝐸−8	−1.53𝐸−9	−1.75𝐸−9	−2.59𝐸−8	−1.20𝐸−8	−0.05
		Re (%) < 120°F	98.55	86.05	97.30	95	78.36	97.69	97.98	98.46

S.D.: Standard deviation.
L: Linear.
NL: Nonlinear.
Re (%) < 120°F: Residual (%, difference between actual and predicted AFTs) less than 120°F.