arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2302.08175cs.ITcs.CVcs.LGmath.IT
收藏

多元正态分布之间 Fisher-Rao 距离的一种数值近似方法

A numerical approximation method for the Fisher-Rao distance between multivariate normal distributions

Sony Computer Science Laboratories Inc(索尼计算机科学实验室公司)

Frank Nielsen

AI总结:

本文通过离散化正态分布间曲线并用 Jeffreys 散度平方根近似局部 Rao 距离,比较多种参数化插值曲线与 Calvo-Oller 等距嵌入曲线,结合上下界评估近似质量。

摘要:

我们提出一种简单方法来近似多元正态分布之间的 Rao 距离:该方法基于离散化连接各正态分布的曲线,并用 Jeffreys 散度(即对称化的 Kullback-Leibler 散度)的平方根来近似曲线上相继相邻正态分布之间的 Rao 距离。我们通过实验考察了正态分布在普通参数化、自然参数化和期望参数化下的线性插值曲线,并将这些曲线与由 Calvo 和 Oller 将 Fisher-Rao d 元正态流形等距嵌入到 (d+1)×(d+1) 对称正定矩阵锥中所得到的曲线进行比较[《Journal of Multivariate Analysis》35.2 (1990): 223-242]。我们报告了实验结果,并通过将数值近似值与下界和上界进行比较来评估该近似技术的质量。最后,我们给出了 Calvo 和 Oller 等距嵌入的若干信息几何性质。

英文摘要:

We present a simple method to approximate Rao's distance between multivariate normal distributions based on discretizing curves joining normal distributions and approximating Rao's distances between successive nearby normal distributions on the curves by the square root of Jeffreys divergence, the symmetrized Kullback-Leibler divergence. We consider experimentally the linear interpolation curves in the ordinary, natural and expectation parameterizations of the normal distributions, and compare these curves with a curve derived from the Calvo and Oller's isometric embedding of the Fisher-Rao $d$-variate normal manifold into the cone of $(d+1)\times (d+1)$ symmetric positive-definite matrices [Journal of multivariate analysis 35.2 (1990): 223-242]. We report on our experiments and assess the quality of our approximation technique by comparing the numerical approximations with both lower and upper bounds. Finally, we present several information-geometric properties of the Calvo and Oller's isometric embedding.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑