Considering the parameter setting of Multi-class Cluster Support Vector Machine ( MCSVM ) lacks scientific basis and parameter selection is much more random , the cuckoo search algorithm is used to improve the MCSVM , and the improved MCSVM is applied to technical status evaluation of armored e-quipment.Finally, through comparing with Particle Swarm Optimization (PSO) MCSVM and Genetic Al-gorithm ( GA) MCSVM, the feasibility and rationality of cuckoo search MCSVM are verified in the tech-nical status evaluation of armored equipment .%针对多簇支持向量机参数设置缺乏科学依据、参数选取随意性较大的问题,利用布谷鸟搜索算法对其进行了改进,并应用改进的多簇支持向量机对装甲装备技术状况进行了评估,最后与粒子群优化算法、遗传算法改进的多簇支持向量机评估结果进行了对比分析,验证了该方法在装甲装备技术状况评估中的可行性与合理性。
展开▼