Iris Data Classification Using Modified Fuzzy C Means

摘要

In the field of real environment, there is a big challenge of clubbing the data according to their working techniques or behavior. Fuzzy clustering techniques can be used where the data belongs to more than one class or bucket decided based on no. of features. It means the decision to classify them in any bucket is to be done by applying some similarity measurements. According to this the data points of any data set can belong to more than one class, even having different membership function value corresponding to different class. Fuzzy clustering technique is comprising of two very dissimilar data types as fuzzy data and usual (crisp) data. It is a kind of function working on probabilistic mode of evaluating the values. Where the whole process is done without training of values to that system i.e. using unsupervised model. In this paper we proposed the modified fuzzy c-means using iris flower data to be clustered followed by proper usage of validation techniques classification entropy(CE) and partition coefficient(PC).