基于混合梯度下降算法的支持向量机参数优化

在解决有限样本、非线性和高维度问题中，建立在统计学习理论和结构风险最小化原则上的支持向量机表现出良好的性能和独特的优势。然而，支持向量机参数对模型的识别精度和泛化能力有很大影响，基于这样的事实，运用混合梯度下降算法对支持向量机的参数进行优化选择，结合量子遗传算法出色的全局优化能力和梯度下降法局部寻优能力，较好地解决了传统的支持向量机中参数优化的难题。最后，用实例验证了该算法的有效性。

Parameter Optimization of SVM Based on HGD

SVM （ Support Vector Machine） based on statistical learning theory and structural risk minimization principle has shown good performance and unique advantages in resolving the non -linear and high dimension problems with limited samples. However, the pa- rameters of SVM have a significant impact on the identifying accuracy and generalization ability of SVM. It is based on the fact that this paper uses HGD （ Hybrid Gradient Descent） ） to optimize the parameters of SVM. HGD combined with the excellent global optimiza- tion capability of QGA （ Quantum Genetic Algorithms） and the excellent local optimization ability of GD （ Gradient Descent） gives a better solution to the traditional problem on Parameter optimization of SVM. Final example demonstrates this algorithm very well.