Does noise affect housing prices? A case study in the urban area of Thessaloniki

EPJ Data Science

Table 7 Results for radius set to 50 m using 5-fold cross-validation for areas A, B and C. The results are presented in terms of the mean absolute error (MAE) and mean absolute percentage error (MAPE). Bold text marks the best score across all models for a given area. The dagger symbol indicates that noise pollution was included in the experiment. The “Noise” column refers to the different noise characteristics: one feature for the average day noise and one for the average night noise (I), one feature which averages both day and night noise (II), one feature for the average day noise (III), one feature for the average night noise (IV) and no features for noise in the baseline model (-)

Model	A			B			C
Model	MAE	MAPE	Noise	MAE	MAPE	Noise	MAE	MAPE	Noise
XGBoost	28,919	0.223	-	22,504	0.15	-	31,511	0.138	-
XGBoost †	28,001	0.229	I	20,858	0.15	I	31,370	0.132	III
LGBM	32,572	0.258	-	21,618	0.158	-	32,752	0.151	-
LGBM †	30,216	0.241	III	22,408	0.161	IV	29,872	0.13	III
RF	31,785	0.256	-	24,224	0.182	-	38,886	0.165	-
RF †	31,380	0.254	II	24,028	0.183	I	39,626	0.168	IV
DT	35,319	0.279	-	33,561	0.27	-	48,802	0.209	-
DT †	35,453	0.271	III	27,756	0.191	III	50,290	0.209	II