A hybrid adaptive decision system for supply chain reconfiguration
提出一种结合智能体仿真与决策树学习的方法，分析手机供应链中运营单元的交互与绩效影响因素，生成运营规则以支持重构决策。
Due to short product life cycle, it is expedient to reconfiguration an existing supply chain from time to time. Companies need to impose the standards on operational units for finding the best or the near best alternative configuration. Thus, it becomes imperative to effectively adapt various enablers in a supply chain by understanding the dynamics between them that help to reconfigure a supply chain for high levels of performance. This paper presents an integration of agent-based simulation and decision tree learning as the data mining techniques to determine adaptive decisions of operational units of a mobile phone supply chain. Agent-based simulation output is subjected to data mining analysis to understand system behaviour in terms of interactions and the factors influencing the performance. An entropy-based formulation is proposed as the basis for comparing different operational units in the supply chain. The insights obtained are then encapsulated as operational rules and guidelines supporting better decision-making.