| | Ref. | Techniques used | Location | Independent variables | Performance evaluation |
| | [13] | ANN with BP neural network | China | Wind velocity and wind direction | Error | | [14] | ANN with BP neural network | — | Wind speed, maximum, minimum, and average temperature, average humidity, air pressure, and rainfall | MAPE and RMSE | | [15] | Recursive least squares algorithm | Sri Lanka | Wind speed | RMSE | | [16] | SVM, NN), RF and k-NN | Gansu province, China | Temperature, wind speed, wind direction, cloud cover, pressure, heat flux, radiation, precipitation, humidity, etc. | MAE, MAPE, RMSE | | [17] | Wavelet method and the improved time series method | China | Wind speed | MAE, MAPE, MSE | | [18] | Multilayer perceptron network and integrated k-NN | — | Wind speed, wind direction, air density, temperature difference, sensible heat flux, and vegetation | MAE and standard deviation of absolute error | | [19] | ANN | Portugal | Wind speed | MAE, RMSE and mean relative error | | [21] | Neural networks and fuzzy logic techniques | Aalborg, Denmark | Wind speed, wind direction, temperature | Normalize mean absolute error, normalized root mean square error | | [22] | ANN | Tamil Nadu, India | Wind speed, relative humidity, and generation hours | RMSE and MAE | | [23] | ANN with BP neural network | Rajasthan, India | Generation hours, relative humidity, and wind speed | MSE and MAE | | [25] | ANN and radial basis function neural networks | Canada | Temperature, dew point temperature, relative humidity, wind direction, wind speed, pressure | MSE, absolute mean error, MAPE, and correlation coefficient |
|
|
