According to the characteristic of SVDD, the proposed algorithm utilizes the non-Gaussian to measure how kernel samples approximate to a spherical area, and then optimize the kernel parameter.
该算法根据支持向量域数据描述本身的特点，利用非高斯性来测量核空间样本接近球形区域分布的程度，并根据此测量结果来优化核参数。
The non-gaussian noise origined from the danamical mechanism , so electromigration danamical information can obtained from non-gaussian analysis of the danamical parameter .
这种非高斯性产生自迁移动力学机制，因此通过非高斯性分析，可以从噪声中提取电迁移相关动力学信息。
应用推荐