基于改进的共轭梯度算法实现的最小二乘隐空间支持向量机
The Implementation of Least Squares Hidden Space Support Vector Machines Based on Improved Conjugate Gradient Algorithm
-
摘要: 本文研究了最小二乘隐空间支持向量机的优化问题.文中采用基于对称超松弛预处理技术改进共轭梯度算法,改进的共轭梯度算法只需求解一个阶数为l-1的线性代数方程组即可,大大节省了计算时间.最后将其应用于最小二乘隐空间支持向量机中建立数学模型,并通过实例验证了该算法的优越性.Abstract: This research main focused on the improvement on Least Squares Hidden Space Support Vector Machines. Using Symmetric Successive Over-Relaxation technology to improve Conjugate Gradient Algorithm, it saves a lot of execution time costs for solving the system of linear algebraic equations with L-1 orders. At the end of this paper, the advantage of this improved algorithm is proved by applying it on mathematics modeling.
下载:
