Spatial distribution of solar PV deployment: an application of the region-based convolutional neural network

EPJ Data Science

Table 1 Model Performance Comparison

Model	Dataset	Algorithm	MAE	RMSE	\(R^{2}\)
M1	PV count per HH	XGBoost	0.038	0.003	62.2%
M2		CATBoost	0.039	0.003	57.4%
M3		LightGBM	0.038	0.003	60.4%
M4		Random Forest	0.039	0.003	60.9%
M5	PV count per HH with energy policy predictors	XGBoost	0.038	0.003	68.5%
M6		CATBoost	0.008	0.0001	66.0%
M7		LightGBM	0.008	0.0001	66.8%
M8		Random Forest	0.009	0.001	61.8%
M9	PV-to-roof ratio	XGBoost	0.009	0.0002	55.7%
M10		CATBoost	0.009	0.0002	56.0%
M11		LightGBM	0.009	0.0001	59.2%
M12		Random Forest	0.009	0.0002	56.0%
M13	PV-to-roof ratio with energy policy predictors	XGBoost	0.008	0.0001	71.1%
M14		CATBoost	0.008	0.0001	66.0%
M15		LightGBM	0.008	0.0001	66.0%
M16		Random Forest	0.009	0.0001	61.8%