An Improved Generalized-Trend-Diffusion-Based Data Imputation for Steel Industry

Table 4

Imputation accuracies with different methods (BFG output).

Missing number	Methods	Group 1			Group 2			Group 3
		RMSE	NRMSE	MAPE	RMSE	NRMSE	MAPE	RMSE	NRMSE	MAPE
	Regression	108.17	0.0598	21.94	97.04	0.0476	17.11	109.20	0.0537	16.60
	EM	87.43	0.0483	18.04	83.43	0.0409	15.38	71.65	0.0352	10.72
3 points	SPLINE	92.99	0.0514	17.55	86.74	0.0437	15.41	129.67	0.0638	17.59
	GTD	75.21	0.0416	16.06	70.73	0.0347	12.50	56.69	0.0279	9.18
	iGTD	69.70	0.0385	14.74	68.64	0.0336	12.26	55.48	0.0273	9.32
	Regression	98.25	0.0457	17.99	93.09	0.0394	15.45	95.38	0.0408	15.49
	EM	77.94	0.0363	14.70	75.53	0.0319	13.09	73.74	0.0315	11.57
4 points	SPLINE	128.22	0.0597	20.51	85.83	0.0363	14.29	66.95	0.0286	10.86
	GTD	75.70	0.0352	14.21	63.36	0.0268	10.47	62.31	0.0266	9.44
	iGTD	70.91	0.0330	13.15	63.25	0.0267	10.32	41.51	0.0178	6.74
	Regression	88.76	0.0362	17.09	79.13	0.0296	11.22	76.24	0.0289	11.47
	EM	72.61	0.0296	12.98	71.74	0.0268	11.98	71.39	0.0270	10.97
5 points	SPLINE	97.83	0.0399	16.72	105.18	0.0393	16.78	102.59	0.0389	15.60
	GTD	67.44	0.0275	11.90	61.00	0.0228	9.63	67.23	0.0255	10.47
	iGTD	67.19	0.0274	11.95	57.75	0.0216	9.29	62.95	0.0238	10.14