论文部分内容阅读
支持向量回归技术广泛用于解决单输出回归问题,但现实中存在更多的是多输出的情形。为更好地解决多输出回归问题,在单输出支持向量回归机的基础上,通过并行运算推广得到一种多输出支持向量回归机,并在动态交通均衡问题的背景下,求解依赖时间的变分不等式问题。实验表明与单输出支持向量回归算法和线性插值比较,多输出支持向量回归算法具有更快的计算速度和更好的拟合效果。文中给出的多输出支持向量回归机不仅推进了多输出支持向量回归机的研究,而且为解决依赖时间的变分不等式问题提供了一种新思路。
Support vector regression is widely used to solve the single-output regression problem, but in reality, there are more cases of multiple output. In order to solve the multi-output regression problem better, a multi-output support vector regression machine based on the single-output support vector regression machine is derived through parallel operation. In the context of dynamic traffic equilibrium problem, Sub-inequality problem. Experiments show that compared with single output support vector regression and linear interpolation, the multiple output support vector regression algorithm has faster computation speed and better fitting effect. The multiple output support vector regression machine presented in this paper not only advances the research of multiple output support vector regression machine, but also provides a new idea for solving the variational inequality dependent on time.