Elastic-Net Regression Algorithm Based on Multi-Scale Gaussian Kernel
Elastic-Net Regression Algorithm Based on Multi-Scale Gaussian Kernel
复制标题
DOI：
--
复制
发表时间：
2014-02
影响因子：
3
通讯作者：
Yong-Li Xu;Zhenjun Yang
中科院分区：
文献类型：
--
作者：
Yong-Li Xu;Zhenjun Yang
提出了一种基于多尺度高斯核的弹性网络算法来处理回归函数的逼近问题。我们使用不同核宽度的高斯核来近似回归函数的高频和低频分量，然后加权预测函数的L1范数和L2范数作为正则化项。在仿真实验中，基于多高斯核的弹性网络比基于单高斯核的弹性网络具有更小的预测误差和更好的稀疏性能。此外，基于多高斯核的弹性网络可以精确地预测目标函数的高频和低频分量。
This paper proposes an elastic-net algorithm based on multi-scale Gaussian kernels to deal with the approximation of regression function. We use Gaussian kernels with different kernel width to approximate the high and low frequency components of the regression function; then weighted L1 norm and L2 norm of the prediction function utilized as regularization term. In the simulation experiments, multiple Gaussian kernels based elastic-net obtains less prediction error and better sparse performance than single Gaussian kernel based elastic-net. In addition, multiple Gaussian kernels based elastic-net can precisely predict the high and low frequency components of objective function.