K-means based approach for categorical and non categorical data sets

Nowadays, corporations and enterprises are used data mining to increase sales and profits through reaching data. Therefore new algorithms and methods are developed in data minig. The machine learning is indispensable component of data mining. In machine learning, there are a lot of algorithms for classification, clustering etc. One of the well known algorithm is K-Means algorithm in machine learning. In K-Means algorithm non categorical data sets are clustering. However in real world categorical and non categorical data sets are nested. The aim of thesis is to develop K-Means algorithm which does clusters categorical and non categorical data sets together. To do this, Jaccard similarity measure is embeded inside K-Means algorithm instead of Euclid for categorical part of data sets then two algorithms are combined each other clustering categorical and non categorical data sets. Key Words: K-Means, Jaccard Similarity Measure, categorical data, non categorical data