Using kmeans(), you can cluster samples according to the specified cluster number.

Currently, the cluster number in SPL only supports 2.

For example, there is a set of samples [[1,2,3,4],[2,3,1,2],[1,1,1,-1],[1,0,-2,-6]], cluster the samples with the cluster number of 2. And use the clustering model to predicting on samples [[6,2,3,5],[0,3,1,5],[1,2,1,-1],[1,5,2,-6]]

A1 Input the training samples

A2 Input the prediction samples

A3 With k=2 as the parameter, a clustering model is established on the sample A1, and return the model information R

A4 Model R in A3 is used to predict on A2 samples and return the prediction result. Samples 1, 2 and 3 are of the same class, and sample 4 is the other class

A5 Continuous modeling and prediction, directly return the prediction results, the effect is equivalent to A3+A4