Open Access Article

Title: Logistics performance index estimating with artificial intelligence

Authors: Bilal Babayigit; Feyza Gürbüz; Berrin Denizhan

Addresses: Department of Computer Engineering, Erciyes University, 38039, Kayseri, Turkey ' Department of Industrial Engineering, Erciyes University, 38039, Kayseri, Turkey ' Department of Industrial Engineering, Sakarya University, 54050, Sakarya, Turkey

Abstract: The World Bank has presented the logistics performance index (LPI) to measure and rank countries' international logistics performance. Based on six different components, the impact of each LPI component should be further investigated. In this paper, performance criteria are ranked using MGGP. This ranking approach is the first kind of study that enables countries to prioritise and adjust measures to evaluate their logistics performance better. MGGP is a recent promising approach among machine learning techniques, and it is capable of creating linear or nonlinear prediction models. LPI datasets consisting of 790 records collected between 2010-2018 are used to train and test the proposed MGGP approach. MGGP help address the logistics performance based on the relative importance of factors. The simulation results show the superiority of the MGGP approach predicting the LPI score. The prediction equation generated by MGGP can be helpful, for policymakers and researchers in logistics, in establishing logistics plans.

Keywords: logistics; logistics performance index; LPI; multi-gene genetic programming; MGGP; artificial intelligence.

DOI: 10.1504/IJSTL.2023.129876

International Journal of Shipping and Transport Logistics, 2023 Vol.16 No.3/4, pp.360 - 371

Received: 31 Jan 2021
Accepted: 09 Nov 2021
Published online: 03 Apr 2023 *

 

Return to top