Showing papers on "Ward's method published in 2021"

Clustering of modal-valued symbolic data

Abstract: Symbolic data analysis is based on special descriptions of data known as symbolic objects (SOs). Such descriptions preserve more detailed information about units and their clusters than the usual representations with mean values. A special type of SO is a representation with frequency or probability distributions (modal values). This representation enables us to simultaneously consider variables of all measurement types during the clustering process. In this paper, we present the theoretical basis for compatible leaders and agglomerative clustering methods with alternative dissimilarities for modal-valued SOs. The leaders method efficiently solves clustering problems with large numbers of units, while the agglomerative method can be applied either alone to a small data set, or to leaders, obtained from the compatible leaders clustering method. We focus on (a) the inclusion of weights that enables clustering representatives to retain the same structure as if clustering only first order units and (b) the selection of relative dissimilarities that produce more interpretable, i.e., meaningful optimal clustering representatives. The usefulness of the proposed methods with adaptations was assessed and substantiated by carefully constructed simulation settings and demonstrated on three different real-world data sets gaining in interpretability from the use of weights (population pyramids and ESS data) or relative dissimilarity (US patents data).

Supplier segmentation : a case study of Mozambican cassava farmers

Abstract: Over 3 000 Mozambican smallholder farmers supply cassava to Company XYZ. XYZ needs an effective supplier segmentation method to gain insight into how it should direct its resources for the greatest impact. This paper describes the application of the k -means algorithm, agglomerative hierarchical clustering, and a self-organising map with ward clustering to segment these cassava suppliers. The insights gained from the cluster analysis are then used to provide recommendations and suggest suitable intervention strategies to manage each segment of suppliers. The proposed method offers users the basis of a supplier segmentation system that is more robust than commonly used qualitative supplier segmentation models.

PRFM Model Developed for the Separation of Enterprise Customers Based on the Distribution Companies of Various Goods and Services

Abstract: In this study, a new model of combining variables affecting the classification of customers is introduced which is based on a distribution system of goods and services. Given the problems that the RFM model has in various distribution systems, a new model for resolving these problems is presented. The core of this model is the older RFM. The new model that has been proposed as PRFM, consists of four dimensions: Profit margins (P), time period from customer's last purchase (R), Frequency of transactions (F) and the Monetary Value (M). Adding variable (P) makes a huge change in customer clustering and classification systems and makes it more optimized for future planning. For review and approval, the model was implemented in one of the largest and most diversified distribution companies in Iran. Using Ward's clustering, the optimal number of clusters was prepared and entered by hierarchical clustering and based on Euclidian distance customers are clustered and separated. One of the most important results of this study is introducing a new model and resolving the problems of the old RFM model in determining customer's value.