Online Learning with Self-tuned Gaussian Kernels: Good Kernel-initialization by Multiscale Screening
Online Learning with Self-tuned Gaussian Kernels: Good Kernel-initialization by Multiscale Screening
复制标题
DOI：
10.1109/icassp.2019.8683899
复制
发表时间：
2019-05
期刊：
影响因子：
--
通讯作者：
Masa-aki Takizawa;M. Yukawa
中科院分区：
文献类型：
--
作者：
Masa-aki Takizawa;M. Yukawa
我们提出了一种有效的内核参数自适应更新方法：内核系数、尺度和中心。分别采用镜像下降法和平方误差成本函数的最速下降法来更新核尺度和中心。尽管本文考虑的问题是非凸的，但我们通过使用新颖的多重初始化方案来增长字典而无需大幅增加字典大小，从而降低了陷入局部最小值的可能性。通过计算机实验，我们表明所提出的算法在保持较小字典大小的同时具有较高的自适应能力，并且无需对初始内核参数进行详细调整。
We propose an efficient adaptive update method for the kernel parameters: the kernel coefficients, scales and centers. The mirror descent and the steepest descent method for squared error cost function are employed to update the kernel scales and centers, respectively. Although the problem considered in this paper is nonconvex, we reduce the possibility of falling into local minima by using a novel multiple initialization scheme to grow the dictionary without great increases of the dictionary size. Through computer experiments, we show that the proposed algorithm enjoys a high adaptation-capability while maintaining a small dictionary size, without detailed tuning of the initial kernel parameters.