METHODICAL APPROACH TO CLUSTER CONFIGURATIONS OF PRODUCT VARIANTS OF COMPLEX PRODUCT PORTFOLIOS

Abstract Companies are increasingly struggling to manage their complex product portfolios. Since they do not fully understand the complexity, intelligent solutions required. Emerging technologies and tools offer new ways deal with existing problems. With help of clustering, similarities between variants can be identified automatically, complexity systematically reduced. This article aims develop a methodological approach identify correlations in portfolios automatically by using clustering algorithms. The includes systematic cleaning transformation portfolio data. In addition, guide for algorithm selection evaluation results is presented. As last step, analysed visualised. To validate methodical approach, it applied real-world data set commercial vehicle manufacturer usefulness confirmed an expert workshop.