Moreover, the K-means algorithm is applicable in the prediction of software faults [29].
The Quad Tree method was added into the K-means algorithm to find the initial cluster.
It was claimed that the combination of the algorithms will make the prediction perform better.
In the context of prediction, Kusiak and Li applied K-means clustering to predict wind power that could be produced by a wind turbine [30].
Five parameters have been investigated for power prediction.
The power produced is categorised into a few clusters up to 1500 kW for different wind speed.
The variation of wind speed varies the power produced.
The performance of the prediction is then tested by using the mean absolute error (MAE), mean relative error (MRE), standard deviation of MAE and standard deviation of MRE [31].