Categorical-and-numerical-attribute data clustering based on a unified similarity metric without knowing cluster number

Comparative Study of Initial Centroid Based K-Mode Algorithm

“…K-Mode based upon unified similarity metrics algorithm [15] proposed a penalized competitive learning algorithm and these algorithm required some initial value of number of clusters which should be greater than the original value of number of clusters. The resulting clusters are more accurate than the original K-Mode Algorithm.…”

An Improved Kernel Clustering Algorithm for Mixed-Type Data in Network Forensic

“…The algorithm introduced the distribution centroid for representing the prototype of categorical attributes and a new dissimilarity measure taking into account the significance of different categorical attributes. Cheung et al [21] proposed a clustering framework based on the concept of object-cluster similarity and defined a unified similarity metric for both numerical and categorical attributes. Liu [22] put forward a clustering algorithm based on average mutual information.…”

A Simple Yet Fast Clustering Approach for Categorical Data

“…ROCK [10] is a popular technique for categorical clustering, Huang's method [5] is based on k-modes and technique of Ahmad and Dey [13] is a variation of k-means. OCIL [20] is a recently proposed categorical clustering algorithm which does not use a priori knowledge of number of clusters. Table 7 compares the results of our algorithm over Heart Disease dataset with those published in [13].…”

“…The popular algorithms for mixed data have been selected for comparison -SBAC [21], Huang [22], ECOWEB [23]and Ahmad-Dey [13]. OCIL [20] is a recently proposed categorical clustering algorithm which does not use a priori knowledge of number of clusters. Thus, it can be concluded that the clustering performance of the proposed algorithm is better than many popular algorithms.…”