Clustering model for in-N-out products in retail: A case study

© 2018, Curran Associates Inc. All rights reserved.Retail management systems have developed rapidly in recent years. They allows companies to strengthen their systems by using advanced data analytics techniques as well as standardize the business practices and improve customer service. One of the main challenges of the retail trade is the planning of the purchase, distribution and evaluation of the performance of the in-N-out products, also known as uncatalogued or temporary ones. Such products represent a high risk for retailers, but at the same time, they are an opportunity to increase marginal contribution rapidly. In this research, we propose a model for analyzing and clustering such products in order to assist some other processes in the retail management. The proposed algorithm [method] is a k-optimal agglomerative one that is tested using real life data during the development of a retail system software solution. The proposed approach was compared with k-means and DB-scan procedures, showing its superiority in providing a better cohesion and separation of individual points and formed clusters.

Clustering model for in-N-out products in retail: A case study Academic Article in Scopus