为进一步增强进化算法在最小属性约简过程中的全局求解性能,提出了一种基于量子云模型反馈的协同精英属性均衡优势集成约简算法,该算法首先设计一种基于云模型反馈的量子自适应旋转角调整策略,使量子蛙群精英在云模型定性知识和罚因子反馈指导下自适应控制属性搜索空间范围;然后构建一种有限理性区域下协同精英均衡优势属性分解框架,在动态精英演化区域内使参与属性约简的量子蛙群精英在平均权重裕度下协同化达到Nash均衡优势区域;最后量子蛙群精英采用集成化操作机制在各自均衡优势区域内协同提取属性约简子集,从而稳定取得全局最优约简特征集。实验结果表明本算法求解全局最优属性约简集效率、精度和稳定性等具有明显优势,应用到孕龄新生儿脑MRIs电子病历分割,进一步表明该算法具有较强的应用性能。 In order to further enhance the global performance of evolutionary algorithm in the minimum attribute reduction process, an integrated reduction algorithm based on quantum cloud model feedback is proposed. This algorithm first designs a quantum self-feedback algorithm based on cloud model feedback Adaptive rotation angle adjustment strategy to enable the quantum frog herd to adaptively control the scope of attribute search space under the cloud model qualitative knowledge and penalty factor feedback; and then construct a decomposition framework of collaborative elite equilibrium advantage attribute under the limited rationality region, In the region, the elitists of quantum frogs who participate in the attribute reduction converge to the Nash equilibrium predominance region under the average weight margin. Finally, the elitists of the quantum frog herd use the integrated operation mechanism to extract the attribute reduction subsets in their equilibrium regions, Thus obtaining the global optimal reduction feature set stably. The experimental results show that this algorithm has obvious advantages such as efficiency, precision and stability in solving the global optimal attribute reduction set, which is applied to the electronic medical records segmentation of neonatal brain MRIs in gestational age, which further shows that the algorithm has strong application performance.