CAJ | 학술논문

海洋环境中各种气象水文要素对海军武器装备的作战效能影响显著,且影响机理复杂,这大大增加了海军武器装备作战效能的评估难度。针对武器装备作战效能评估问题中的小样本、非线性和高维度等问题,将支持向量回归机( Sup-port Vector Regression,SVR)模型应用到该作战效能评估问题中会有较好的性能表现。标准的支持向量回归机对每个样本公平对待,而实际中的样本存在重要性的差别,文中将模糊支持向量回归机应用到武器作战效能评估问题中,提出一种根据样本与所有测试样本在高维特征空间的欧氏距离之和的大小来确定每个样本模糊隶属度的方法,从而体现不同的样本对决策函数学习的贡献率差异。实验结果表明,对于样本中存在野点和噪声的回归问题,模糊支持向量机表现出了更高的评估精度。
해양배경중각충기상수문요소대해군무기장비적작전효능영향현저,차영향궤리복잡,저대대증가료해군무기장비작전효능적평고난도。침대무기장비작전효능평고문제중적소양본、비선성화고유도등문제,장지지향량회귀궤( Sup-port Vector Regression,SVR)모형응용도해작전효능평고문제중회유교호적성능표현。표준적지지향량회귀궤대매개양본공평대대,이실제중적양본존재중요성적차별,문중장모호지지향량회귀궤응용도무기작전효능평고문제중,제출일충근거양본여소유측시양본재고유특정공간적구씨거리지화적대소래학정매개양본모호대속도적방법,종이체현불동적양본대결책함수학습적공헌솔차이。실험결과표명,대우양본중존재야점화조성적회귀문제,모호지지향량궤표현출료경고적평고정도。
Operational effectiveness of naval weapon equipment is influenced by the meteorological and hydrological elements of marine environment with complex mechanism,which greatly enhances the difficulty of its operational effectiveness evaluation. Aiming at the non-linear and multi-dimension problem with small samples in the weapon equipment operational effectiveness evaluation,Support Vector Re-gression ( SVR) model is applied to the operational effectiveness evaluation of weapon to make good performances. In the theory of SVR,all the training samples are treated uniformly,but in real-world applications,the effects of the training samples are different. In this paper,introduce Fuzzy Support Vector Regression ( FSVR) to problems of operational effectiveness evaluation of weapon and propose a method of giving each sample a fuzzy membership according to the sum of the distances from the sample and all the test samples in the high dimensional feature space,so that different samples can make different contributions to the learning of decision function. Experimen-tal results show that FSVR shows a higher evaluating accuracy in regression problems with outliers or noises in the samples.