Integrated Logistics Performance Assessment Using Machine Learning

Abstract

This study proposes a machine-learning based framework for the construction and interpretation of a composite indicator for logistics performance assessment. The methodology begins with two PRISMA systematic literature reviews conducted to identify and select a relevant portfolio of studies. Based on the resulting theoretical and methodological framework, a Python-based algorithm was developed to apply a Random Forest Regressor for the construction of the composite indicator. The final stage of the study focuses on the analysis of the results generated by the algorithm, including the temporal evolution of the composite indicator and the contribution of individual Key Performance Indicators (KPIs), in order to investigate their potential influence on the performance of the logistics operation.