International Journal of Neutrosophic Science IJNS 2690-6805 2692-6148 10.54216/IJNS https://www.americaspg.com/journals/show/4167 2020 2020 Research Institute of Business Analytics and SCM, College of Management, Shenzhen University, China; Center for Reseaech and Innovation, Asia International University, Yangibod MFY, G’ijduvon street, House 74, Bukhara, Uzbekistan; Faculty of Engineering and Quantity Surviving, INTI International University Colleges, Nilai, Malaysia Muhammad Muhammad Department of Mathematics, Govt College University Faisalabad, Pakistan Anns Uzair Department of Mathematics, Institute of Numerical Sciences, Gomal University, Dera Ismail Khan, 29050, KPK, Pakistan Muhammad Tahir Faculty of Arts and Sciences, Department of Mathematics, Near East University, Nicosia, 99010, Turkey; Jadara University Research Center, Jadara University, Irbid, Jordan; Research Center of Applied Mathematics, Khazar University, Baku, Azerbaijan Muhammad Farman Center for Reseaech and Innovation, Asia International University, Yangibod MFY, G’ijduvon street, House 74, Bukhara, Uzbekistan Ixtiyarov Farxod Faculty of Engineering and Quantity Surviving, INTI International University Colleges, Nilai, Malaysia; Faculty of Mangement, Shinawatra, Pathum Thani, Thailand Mohamed Hafez Transportation optimization remains a critical challenge in international businesses, particularly given the inherent uncertainties of supply chain networks. This paper proposes a novel machine learning-based model for solving multi-objective, multi-item solid transportation problems that fundamentally advances beyond existing fuzzy and neutrosophic approaches. Our key innovation lies in the synergistic integration of neutrosophic Z-numbers (NZNs) with adaptive machine learning techniques, creating a framework that simultaneously captures value vagueness, information reliability, and dynamic uncertainty patterns capabilities absent in conventional fuzzy transportation models. Unlike traditional fuzzy methods that treat all uncertainty uniformly, our NZN representation provides a three-dimensional structure incorporating truth, indeterminacy, and falsity measures, each with associated reliability metrics. This enriched uncertainty modeling enables three ground breaking advancements over existing approaches: (1) a neural scoring system that autonomously learns optimal NZN comparison functions from historical decision patterns, overcoming the limitations of static aggregation operators in fuzzy systems; (2) LSTM networks that jointly forecast demand values and their reliability evolution under uncertainty; and (3) reinforcement learning optimizers that dynamically balance economic efficiency with information quality in routing decisions. Computational experiments demonstrate superior performance compared to six established baseline methods, including traditional fuzzy, intuitionistic fuzzy, neutrosophic, and pure machine learning approaches. Our hybrid framework achieves a 23.4% reduction in transportation costs and 35.4% improvement in uncertainty handling compared to conventional fuzzy transportation models, with statistically significant improvements (p < 0.001) across all evaluation metrics. By coupling the theoretical rigor of neutrosophic mathematics with the adaptive power of machine learning, this study provides businesses with a transformative decision-support system for transportation planning under real-world uncertainty conditions. 2025 2025 226 253 10.54216/IJNS.260421 https://www.americaspg.com/articleinfo/21/show/4167