K-means algorithm

K-meansAlgorithmIt is an interesting data clustering algorithm. The original meaning is to divide n objects into k regions and K <n. When I see that the first idea of this algorithm is whether it is related to the canvas. In fact, it is true that the points on the boundaries of the two clusters generally constitute the broken line segments of the canvas. This algorithm is widely used in molecular biology, and it is very powerful for CG-Cornell's guys to use this algorithm to accelerate the computation of Indirect lighting.