Author(s): R. Aruna Kirithika; K. Padmavathi

The logistics division in any trade organization is a typical task that involves a huge man power, time and needs more currency to be spent while ensuring safe delivery of the product involved. Any organization that does trade across the world is in need to maintain a huge database that has records of almost numerous numbers of attributes involved in it. Either a survey or a thorough study of the database will be a tremendous process. Here in this paper such a logistics division database is traced with the help of the clustering algorithm to show the performance of the clusters with the data given. The response and retrieval of data avoiding the noisy data from the database is a huge task. Such process is done with the help of clustering algorithms. Upon applying the database to various clustering algorithms certain results have been found and discussed. Three clustering algorithms have been chosen for analysis such as K-means, EM and Make density based algorithm are applied with the same database and a comparative results have been derived in this paper. The clustering algorithms used for the prediction of results are based upon the statistical calculation. All algorithms that are used in the field of Data Mining make use of the statistical equation to predict results out of the entire data warehouse for all bounded and arbitrary values.