A k-means clustering for supply chain risk management with embedded network connectivity

Abstract

Abstract:In recent years, increased attention has been shown to the supply chain risk management due to the occurrences of several high profile disruptions which resulted in significant social, economic and political impact globally. However, there are not direct and easy ways of understanding the risk of an entire supply chain. In this paper, a network connectivity embedded k-means clustering approach has been proposed to determine at-risk clusters of nodes that share similar risk profiles and linkages with the focal company. It uses a multiple dimensional feature vector to represent the risks that nodes are facing, their geographical locations, supply chain attributes and network connectivity attributes. The clustering approach is able to reduce the complexity of a large supply chain network to facilitate in-depth targeted analysis and simulations. The effectiveness of the proposed approach has been illustrated by experiments that successfully identify the risk clusters and critical risk zones.

An integrated recommender system for improved accuracy and aggregate diversity

“…They employed a clustering technique on implicit data to reduce the size of the dataset and dimensionality of the item space for improving accuracy. Similarly, weighted clustering and k-means clustering methods were employed for collaborative filtering to improve the quality of recommendations (Kant el al., 2018;Salah et al, 2016;Yin et al, 2016). Kwon (2012, 2014) proposed a rank-based method and an optimization based approach to improve aggregate recommendation diversity.…”