Date of Graduation


Document Type


Degree Name

Doctor of Philosophy in Engineering (PhD)

Degree Level



Industrial Engineering


Manuel D. Rossetti

Committee Member

Edward A. Pohl

Second Committee Member

Nebil Buyurgan

Third Committee Member

Scott J. Mason


Applied sciences, Classification, Clustering, Grouping techniques, Large scale inventory system, Network of item type


Large retail companies operate large-scale systems which may consist of thousands of stores. These retail stores and their suppliers, such as warehouses and manufacturers, form a large-scale multi-item multi-echelon inventory supply network. Operations of this kind of inventory system require a large number of human resources, computing capacity, etc.

In this research, three kinds of grouping techniques are investigated to make the large-scale inventory system “easier” to manage. The first grouping technique is a network based ABC classification method. A new classification criterion is developed so that the inventory network characteristics are included in the classification process, and this criterion is shown to be better than the traditional annual dollar usage criterion. The second grouping technique is “NIT” classification, which takes into consideration the supply structure of the inventory item types. In order to have similar operations-related attributes for items within the same group, a network based K-Means clustering methodology is developed to cluster items based on distance measures. It is believed that there is no single best model or approach to solve the problems of the complex multi-item multi-echelon inventory systems of interest. Therefore, some combinations of different grouping techniques are suggested to handle these problems.

The performance of the grouping techniques are evaluated based on effectiveness (grouping penalty cost and Sum of Squared Error) and efficiency (grouping time). Extensive experiments based on 1,024 different inventory system scenarios are carried out to evaluate the performance of the ABC classification, NIT classification, and the K-Means clustering techniques. Based on these experimental results, the characteristics of the 3 individual grouping techniques are summarized, and their performance compared. Based on the characteristics and performance of these grouping techniques, suggestions are made to select an appropriate grouping method.