Abstract
Traditional clustering algorithm can't deal with non-linear fuzzy and boundary problem. This paper provides a rough fuzzy kernel clustering algorithm. The algorithm firstly using kernel function map input space to high-dimensional space, make the space can be partitioned linearly. Then it using rough and fuzzy theory computes the membership grade difference between sample data and cluster center. It compares the difference and γ threshold, judges the sample data belonging to lower or upper approximation. At last, it computes the center of rough clusters again. Through these iterative computations, it can get the clustering result. This algorithm combines kernel method and rough fuzzy theory, the experiment results show this method is effective.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
