Assessing Green Logistics and Supply Chain Resilience With Future Importance Analysis: Machine Learning and Multicriteria Decision-Making Approach

dc.contributor.authorSimsek, Ahmed Ihsan
dc.contributor.authorKoc, Erdinc
dc.contributor.authorGultekin Tarla, Esma
dc.date.accessioned2026-08-12T17:42:44Z
dc.date.issued2026
dc.departmentFırat Üniversitesi
dc.description.abstractThis study investigates the effectiveness of sustainability-oriented factors in supply chain management and their effects on supply chain resilience. Using the Supply Chain Management with Green Logistics dataset obtained from the Kaggle platform, 19 basic supply chain components of 69 companies were examined with machine learning and multicriteria decision-making (MCDM) methods. The modeling performed using Random Forest, Gradient Boosting, XGBoost, LightGBM, CatBoost, MLPRegressor, Lasso, Ridge, SVR, AdaBoost, and ExtraTrees algorithms was evaluated with performance metrics such as RMSE, MSE, MAE, MAPE, and R 2, and the AdaBoost algorithm showed the best performance. In order to improve the performance of the model, fivefold K-fold cross-validation and hyperparameter optimization with GridSearch were performed. In the feature importance analysis, Order Fulfillment Rate stood out as the variable with the highest impact score, whereas sustainability-oriented variables (recycling rate, carbon emissions and use of renewable energy) were found to be of lower importance. The results obtained from the study show that the most important variable is order fulfillment and customer focus. Within this framework, according to the results obtained for companies that have green-focused processes, traditional supply chain elements are more important. Sensitivity analyses conducted with ADAM, CoCoSo, and MABAC methods examined the effects of changes in the weights of these variables on the results. The findings highlight the limited impact of green logistics practices on the efficiency of enterprises in the short term and show the importance of including these factors in strategic planning processes. This indicates that environmental sustainability should be supported by policy-oriented interventions rather than market mechanisms. In this context, structural policy changes are needed, such as providing tax breaks and appropriate financing opportunities for green logistics investments, as well as encouraging logistics operations with low carbon footprints through certification and providing competitive advantages to these companies.
dc.identifier.doi10.1002/bse.70356
dc.identifier.endpage3977
dc.identifier.issn0964-4733
dc.identifier.issn1099-0836
dc.identifier.issue3
dc.identifier.orcid0000-0002-2900-3032
dc.identifier.orcid0000-0001-5897-0462
dc.identifier.orcid0000-0002-8209-5714
dc.identifier.scopus2-s2.0-105022710793
dc.identifier.scopusqualityQ1
dc.identifier.startpage3952
dc.identifier.urihttps://doi.org/10.1002/bse.70356
dc.identifier.urihttps://hdl.handle.net/11508/59853
dc.identifier.volume35
dc.identifier.wosWOS:001620503500001
dc.identifier.wosqualityQ1
dc.indekslendigikaynakWeb of Science
dc.indekslendigikaynakScopus
dc.language.isoen
dc.publisherWiley
dc.relation.ispartofBusiness Strategy and the Environment
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/closedAccess
dc.snmzKA_WoS_20260511
dc.subjectAdaBoost
dc.subjectgreen logistics
dc.subjectMCDM
dc.subjectsupply chain resilience
dc.subjectsustainability
dc.titleAssessing Green Logistics and Supply Chain Resilience With Future Importance Analysis: Machine Learning and Multicriteria Decision-Making Approach
dc.typeArticle

Dosyalar