Pavel Foltin
University of Defence in Brno, Faculty of Military Leadership, Department of Logistics, Czech Republic
E-mail: pavel.foltin@unob.cz

Martin Brunclík
University of Defence in Brno, Faculty of Military Leadership, Department of Logistics, Czech Republic
E-mail: martin.brunclik.me@gmail.com

Vojtěch Ondryhal
University of Defence in Brno, Faculty of Military Technology, Department of Communication and Information Systems, Czech Republic
E-mail: vojtech.ondryhal@unob.cz

Lukáš Vogal
University of Defence in Brno, Faculty of Military Leadership, Department of Logistics, Czech Republic
E-mail: lukas.vogal@unob.cz

Abstract

Level of supply chain performance is highly influenced by capability of logistics infrastructure as well as created network of activities and services operating at global, national, and regional scales. Dynamic environment can significantly influence expected availability and real capability of logistics infrastructure. The article represents possible analytical approach to supply chain design and identification of factors influencing logistics performance. In addition, the paper deals with application of key indicators and indexes to the classification of individual national levels of available logistics infrastructure. Clustering algorithms (k-means and hierarchical) and classification methods (decision trees, kNN and logistic regression) have been used for classification model construction and results evaluation. Furthermore, possible application for forecasting logistics infrastructure availability has been discussed.

Key words: logistic infrastructure, logistic performance index (LPI), country classification, clustering algorithms

Last Update: October 17, 2018