机构地区: 西华大学数学与计算机学院
出 处: 《西南民族大学学报(自然科学版)》 2006年第5期1007-1011,共5页
摘 要: 使用线性规划优化技术代替二次规划优化技术,通过最小化支持向量数来实现支持向量机算法.由于线性规划支持向量机的核函数不需要满足Mercer定理,因此,采用复高斯小波B样条小波作为支持向量机的核函数,建立了线性规划支持向量机模型,并将其用于非线性系统的辨识.仿真结果表明,线性规划支持向量机模型的辨识精度高于二次规划支持向量机模型. Firstly,we substitute linear programming problem for quadratic programming problem by minimizing the number of support vector machines.Secondly,because the kernel functions are not acquired to satisfy the Mercer theorem,the complex Gauss B-spline wavelet is a kernel for support vector machines and a linear programming support vector machine model is established.Finally,the model is used to identify the non-linear system.The simulation results show that the preciseness of identification of the linear programming technique is higher than the quadratic programming technique.