Facilitating the extraction of extended insights on logistics performance from the logistics performance index dataset: A two-stage methodological framework and its application

摘要: Abstract The Logistics Performance Index (LPI) developed by the World Bank provides a comparative assessment logistics performance in trade for several countries. Given lack of studies bringing insights on backdrop from perspective nation as whole—this paper recognizes LPI dataset an account rich country-level data with harbored performance. It further suggests that upon linking appropriate variable(s) interest, extended can be extracted. Therefore, two-stage methodological framework is suggested mining towards insights. first stage involves clustering into finite clusters using K-means algorithm. Subsequently, second stage, suitability multivariate adaptive regression spline (MARS) based outlined capturing complex non-linear relationship between variables under investigation. Thereby, application proposed demonstrated example six dimensions important macroeconomic variable. In addition to discussing critical implications performance, and its utility framework—the findings suggest extraction governed selection which linked dimensions, criteria development MARS models overall data.