Application of Fuzzy and Conventional Forecasting Techniques to Predict Energy Consumption in Buildings

Table 6

Comparison between fuzzy and nonfuzzy approaches.

Meter	Approach	RMSE	MAE	R^2
M1
	Fuzzy	5.1093	1.6630	0.9245
	Nonfuzzy	5.1306	1.5869	0.9239
M2
	Fuzzy	10.8719	2.5343	0.8524
	Nonfuzzy	10.6464	2.9834	0.8584
M3
	Fuzzy	4.6532	0.9554	0.9340
	Nonfuzzy	4.7567	1.4236	0.9311
M4
	Fuzzy	10.1662	2.3978	0.9857
	Nonfuzzy	10.1513	2.6782	0.9858
M5
	Fuzzy	6.8727	1.6164	0.9757
	Nonfuzzy	6.7014	0.9829	0.9769
M6
	Fuzzy	4.6280	0.8201	0.9617
	Nonfuzzy	4.8017	1.0286	0.9588
M7
	Fuzzy	7.9172	1.3358	0.9448
	Nonfuzzy	7.8352	1.2313	0.9460
M8
	Fuzzy	4.1679	1.1607	0.9228
	Nonfuzzy	4.0838	0.8277	0.9259
M9
	Fuzzy	4.1845	1.6167	0.8948
	Nonfuzzy	4.1723	1.7256	0.8954
M10
	Fuzzy	5.5033	1.3785	0.9571
	Nonfuzzy	5.5571	1.3467	0.9563

Bold values represent the best approach for each metric; that is to say, for each column (metric), we compare the two approaches (fuzzy and nonfuzzy), and the best is highlighted in bold.