论文部分内容阅读
支持向量机属于一种极其学习方法,依据统计学习理论的思路。在模拟生物的启发下,产生了神经网络学习算法的基本构造,同时支持向量机的理论来自于最小化错误率的理论边界。这里所说的支持向量机具有唯一解并且不依赖输入空间的相关维数,有着较好的数学性质。同时在实际应用中具有较好的性能。本文探究了最小二乘支持向量机参数优化的相关问题。
Support vector machine belongs to an extremely learning method, based on the idea of statistical learning theory. Inspired by simulated creatures, the basic construction of neural network learning algorithm is generated, and the theory of SVM comes from the theoretical boundary of minimizing the error rate. The SVM mentioned here has the unique solution and does not depend on the relevant dimension of the input space, and has good mathematical properties. At the same time in the practical application has good performance. This paper explores the issues related to optimization of parameters in least squares support vector machines.