Power Prediction of Combined Cycle Power Plant (CCPP) Using Machine Learning Algorithm-Based Paradigm

Table 5

Effect of feature selection on RMSE and AE for all the algorithms. The table is horizontally partitioned to emphasize on the combinations with and without the temperature (TEMP).

Features combinations	LR		GBRT		KNN		ANN		DNN		Mean RMSE and AE for combinations
	RMSE	AE	RMSE	AE	RMSE	AE	RMSE	AE	RMSE	AE	RMSE-FEATURES	AE-FEATURES
HUM	15.66	13.15	17.31	13.95	18.19	14.72	15.86	13.69	15.64	13.15	16.53	13.73
PRE	14.39	11.67	15.41	11.62	16.5	12.93	15.55	13.24	14.15	11.42	15.2	12.18
PRE-HUM	13.12	10.63	13.72	10.4	14.19	1.01	13.92	11.7	12.82	10.21	13.55	8.79
VAC	8.43	6.56	5.72	3.99	6.16	4.41	7.78	6.05	7.64	6.02	7.15	5.41
VAC-HUM	8.15	6.39	4.63	3.39	7.48	5.5	7.55	5.93	7.36	5.76	7.03	5.39
VAC-PRE	7.85	6.14	4.45	3.11	6.05	4.24	7.1	5.49	6.93	5.29	6.48	4.85
VAC-PRE-HUM	7.52	5.85	4.15	2.86	5.75	4.08	6.72	5.21	6.35	4.79	6.1	4.56
TEMP	5.07	4.11	5.1	4.03	5.51	4.41	4.89	3.91	4.7	3.75	5.05	4.04
TEMP-PRE	5.01	4.09	4.9	3.71	5.15	3.93	4.67	3.77	4.53	3.63	4.85	3.83
TEMP-HUM	4.43	3.55	4.45	3.5	4.69	3.74	4.39	3.43	4.12	3.25	4.42	3.49
TEMP-PRE-HUM	4.43	3.55	4.01	3.01	4.2	3.12	4.31	3.39	4.06	3.16	4.2	3.25
TEMP-VAC	4.7	3.77	2.9	2.22	4.12	3.09	4.41	3.52	4.2	3.35	4.07	3.19
TEMP-VAC-HUM	4.27	3.4	2.82	2.11	3.79	2.86	4.19	3.33	3.84	3.01	3.78	2.94
TEMP-VAC-PRE	4.63	3.7	2.63	1.96	3.57	2.53	4.16	3.32	3.88	2.96	3.77	2.89
TEMP-VAC-PRE-HUM	4.26	3.4	2.64	1.93	3.32	2.37	4.16	3.34	3.61	2.82	3.6	2.77