Information bunching is a typical system for measurable information investigation. Grouping gives apportioning of an informational index into subsets of comparable items or information bunches. Quality bunching is the way toward gathering related qualities in a similar group. The set of fuzzy is unmistakable in a fresh set; it enables those components which contain a level of participation esteem. The center of a set of fuzzy data is its enrolment work: a capacity that characterizes the connection among an incentive and their level of participation in the set of fuzzy data. The Fuzzy C-Means approach of bunching experiences a few compels that influence the execution of information. That is, the enrolment of an information indicate in a bunch depends straightforwardly on its participation esteems in other group focuses and this occasionally happen to create doubtful outcomes. To beat these downsides, the new calculation called Covering Rough Set Fuzzy C Medoids (CRFM) has been proposed. Covering Rough set model requires reducible component and neighbourhood work. In light of this capacity, centroids are figured and that will be considered as a contribution to the Fuzzy C Means calculation for grouping. The proposed calculation is a combination of Covering Rough data set and Fuzzy set calculation. For examination, the proposed calculation is contrasted and FCM and Weighted K Means (WKM) calculation and tried against cerebrum tumour quality articulation dataset. The execution of these calculations result demonstrates that our proposed CRFM calculation furnishes the best groups with enhanced effectiveness contrasted with other two bunching systems. © 2017, Institute of Advanced Scientific Research, Inc. All rights reserved.