Wind speed prediction for site selection and reliable operation of wind power plants in coastal regions using machine learning algorithm variants

Sustainable Energy Research

Table 14 Quartile percent of the prediction error for Dataset 2 (minimum Std. deviation and IQR are bolded)

Model	Quartile percentile for Dataset 2: Kutubdia					Quartile percentile for Dataset 2: Cox's Bazar
Model	Std	25%	50%	IQR	75%	Std	25%	50%	IQR	75%
MLR	0.7008	− 0.4395	− 0.0139	0.8436	0.4041	0.7372	− 0.4412	− 0.0076	0.8657	0.4245
Lasso	0.7076	− 0.4460	− 0.0233	0.8483	0.4023	0.7475	− 0.4609	− 0.0149	0.8856	0.4247
Ridge	0.7008	− 0.4397	− 0.0138	0.8439	0.4042	0.7372	− 0.4413	− 0.0076	0.8658	0.4245
Elastic Net	0.7062	− 0.4522	− 0.0226	0.8513	0.3992	0.7471	− 0.4636	− 0.0143	0.8886	0.4250
KNN	0.6226	− 0.3589	− 0.0039	0.7144	0.3556	0.6634	− 0.3722	0.0039	0.7433	0.3711
DT	0.6740	− 0.4009	− 0.0105	0.7798	0.3789	0.7052	− 0.4224	− 0.0109	0.8216	0.3992
RF	0.6433	− 0.3806	− 0.0050	0.7451	0.3645	0.6672	− 0.4013	− 0.0085	0.7703	0.3690
GBR	0.5871	− 0.3304	− 0.0112	0.6538	0.3234	0.6191	− 0.3473	0.0003	0.6915	0.3442
AdaBoost	0.7181	− 0.4642	− 0.0178	0.8704	0.4061	0.7544	− 0.4810	− 0.0311	0.8953	0.4143
XGBoost	0.5856	− 0.3321	− 0.0104	0.6529	0.3208	0.6199	− 0.3510	− 0.0034	0.6966	0.3456
LightGBM	0.5871	− 0.3327	− 0.0109	0.6521	0.3194	0.6192	− 0.3464	− 0.0026	0.6899	0.3436
CatBoost	0.5749	− 0.3198	− 0.0052	0.6369	0.3171	0.6119	− 0.3374	− 0.0007	0.6730	0.3356
LSTM	0.6088	− 0.2954	0.0561	0.7048	0.4095	0.6580	− 0.3558	0.0254	0.7682	0.4124
GRU	0.6078	− 0.3200	0.0394	0.7097	0.3897	0.6757	− 0.3779	0.0311	0.8108	0.4329