Randomly select k data objects from the data set n as the initial clustering center, and find the distance between other data and this k number, classify according to the principle of minimum distance, and then average each class to determine Central point. Repeat the calculation until the end condition is met.