Visualization and data mining method for inventory classification

Abstract

Inventory classification of stock-keeping units is typically achieved with ABC analysis accompanied with a diagram describing the distribution of dollar-usage values. Unfortunately, it fails to identify the interdependencies between the products, which may lead to alienating customers by ignoring the effect of assortments of choices. To remedy this problem, we propose new visualization and product interdependency identification methods. Experimental results in real-world scenarios for two warehouse datasets are included and analyzed.