NOVEL CLUSTERING METHOD FOR THE CATEGORICAL DATA USING MATHEMATICAL FUZZY PARTITIONING

Abstract

Past subtractive clustering techniques can be utilized for numerical information, yet it can't be connected to downright information since trait estimations of clear cut information don't have a characteristic requesting. The k-modes clustering calculation is without a doubt a standout amongst the most broadly utilized partitioned calculations for straight out information. By utilizing a basic coordinating difference measure for all out articles and modes rather than means for cluster, another approach is created, which permits the utilization of the k-implies worldview to productively group huge clear cut informational collections. A fluffy k-modes calculation is introduced and the viability of the calculation is shown with exploratory outcomes.