Niching Genetic Programming for Co-Joint Optimization: A Case Study on Replenishment and Transshipment Policies in Dynamic Multi-Site Inventory Management

Meng Xu, Jinsheng Chen, Doan Ngoc Chi Nam, Joel Wei En Tay, Puay Siew Tan, Yew-Soon Ong

Abstract: This paper proposes a novel niching genetic programming (GP) framework for co-joint optimization, designed to maintain exploration and solution quality in dynamic environments. The framework comprises a multi-tree representation, with separate trees encoding the respective optimization tasks, and incorporates a niching mechanism to preserve diversity and prevent premature convergence. This integrated approach enables comprehensive exploration of the search space and the discovery of effective, synergistic policies. We apply this framework to the joint optimization of replenishment and transshipment policies in dynamic multi-site inventory management, a critical and challenging problem where policy interdependence greatly expands the search space. Traditional methods, including classical GP, struggle to adapt to rapidly changing environments or fail to capture such co-joint dependencies. Experimental results on synthetic and real-world datasets demonstrate the superiority of the proposed framework over state-of-the-art methods, offering a scalable and robust solution for co-joint optimization in dynamic systems.

External IDs:dblp:journals/memetic/XuCNTTO26