From Table17, the average wind speed for the observed and Mycielski-3 algorithm matched perfectly. However, the line graph shown in Fig. 9 indicates that the Mycielski-3 prediction over predicted the wind speed in June and October. The larger gap in the prediction of the Mycielski-3 of these 2 months compensated for the under-predicted wind speed in January. Meanwhile, K-means clustering does not consist of anyover-predictions that could compensate for the error in the first month. Although K-means clustering does not match the average wind speed, the RMSE value is significantly smaller than the Mycielski-3predic- tion. Hence, K-means clustering prediction is preferable for this research as the RMSE of the method is the smallest. Additionally, the wind profile of K-means clustering is more similar to the observed wind speed when compared with the Mycielski-3 algorithm.