论文部分内容阅读
基于最优化理论中的KKT互补条件建立支持向量回归机的无约束不可微优化模型,并给出了一种有效的光滑近似解法——调节熵函数方法.该方法不需参数取值很大便可逼近问题的最优解,从而避免了一般熵函数法为了逼近精确解,参数取得过大而导致数值的溢出现象,为求解支持向量回归机提供了一条新途径.数值实验结果表明,回归型支持向量机的调节熵函数法改善了支持向量机的回归性能和效率.
Based on the KKT complementation condition in the optimization theory, an unconstrained and non-differentiable optimization model of support vector regression machine is established, and an effective smoothing approximation method - adjusting entropy function method is given. The optimal solution of the problem is approximated so as to avoid the overflow phenomenon of the general entropy function method in order to approximate the exact solution and get the parameter too large, which provides a new way to solve the support vector regression.Numerical experimental results show that regression- Regulated entropy function of vector machine improves the regression performance and efficiency of support vector machine.