Comparative Performance Evaluation of Clustering Algorithms for Grouping Manufacturing Firms

Abstract

Surviving and remaining profitable in market is the biggest requirement for any enterprise in today s business environment because of the stiff competition. Being innovative and staying updated with the latest and trending changes happening in the sector is the key for firms to leave their impression as successful and profit building business house. However, it is not necessary that all the firms should follow same pattern of structure or production procedure even if they come under the same cap of industry but at some point of time up gradation is needed. Appropriate grouping of various manufacturing firms plays a vital role in assessing their credentials. Data relating to appropriate attributes of three types of manufacturing firms are collected. By employing five different clustering techniques each type of these firms are grouped and rank of each method and for each category of industry is evaluated and presented. Results obtained from analysis demonstrate that the overall ranking based on cluster potentiality of different methods is ordered as Self Organized Maps, Gaussian Mixture Model, Fuzzy C-Means, K-Means and Hierarchical techniques. Finding of this study can help the decision makers to devise appropriate strategy for their production pattern according to the firm capability. 2017, King Fahd University of Petroleum & Minerals.