Research Article

Short-Term Wind Speed Forecasting Using Support Vector Regression Optimized by Cuckoo Optimization Algorithm

Table 3

Performance evaluation for one-step- to four-step-ahead predictions.

PredictionsMethodsMAE (m/s)RMSE (m/s)MAPE (%) (—)IA (—)

One-step-ahead predictionSVR0.6911 0.9300 9.6650 0.9424 0.9846
GA-SVR0.6860 0.9184 9.6524 0.9435 0.9851
PSO-SVR0.6951 0.9376 9.6896 0.9414 0.9844
COA-SVR0.6836 0.9119 9.6371 0.9443 0.9853

Two-step-ahead predictionSVR1.0101 1.3598 14.6636 0.8779 0.9657
GA-SVR1.0418 1.4028 14.7376 0.8706 0.9630
PSO-SVR1.0472 1.4148 14.7800 0.8697 0.9618
COA-SVR1.0051 1.3420 14.6764 0.8800 0.9673

Three-step-ahead predictionSVR1.1608 1.5941 17.3654 0.8318 0.9521
GA-SVR1.1814 1.6129 17.5094 0.8285 0.9502
PSO-SVR1.2290 1.6662 17.7517 0.8181 0.9460
COA-SVR1.1515 1.5815 17.3018 0.8343 0.9539

Four-step-ahead predictionSVR1.2602 1.7247 18.9360 0.8043 0.9423
GA-SVR1.2616 1.7293 18.9662 0.8032 0.9421
PSO-SVR1.2761 1.7502 19.0296 0.7999 0.9398
COA-SVR1.2386 1.7079 18.8183 0.8075 0.9451