Abstract: The traditional K-means algorithm often leads to unstable clustering quality due to the randomness of the initial clustering center selection and tends to fall into suboptimal solutions when ...