Identification of supply chain disruptions with economic performance of firms using multi-category support vector machines

Abstract

From a practical perspective, a novel method using multi-category support vector machines (MC-SVM) is proposed to identify supply chain disruptions (SCD). With the data related to economic performance from quarter statements and individual announcements published by the listed firms, the variables of MC-SVM are constructed firstly. Secondly, MC-SVM is used for matching the portfolios of firms, which helps the map from economic performance to SCD by applying MC-SVM again. Finally, a case study is given to testify the ability of proposed method with the data from the listed firms in China.

Keywords:

• identification
• supply chain disruptions
• multi-category support vector machines
• economic performance of firms
• matching portfolios

Acknowledgements

The data of economic performance is from the quarter statements and the individual announcements published by listed firms in China from 2007 to 2011, and it is available on the web page of the Data Centre of Finance Sina Online (http://finance.sina.com.cn), the Securities Times (http://www.stcn.com) and the Shanghai Securities News (http://english.cnstock.com). The training and simulating of MC-SVM is performed with the LS-SVMlab Toolbox developed by K. De Brabanter, et al., Katholieke Universiteit Leuven (http://www.esat.kuleuven.be/sista/lssvmlab).