Vardi Y, Zhang C H
Department of Statistics, Rutgers University, New Brunswick, NJ 08854, USA.
Proc Natl Acad Sci U S A. 2000 Feb 15;97(4):1423-6. doi: 10.1073/pnas.97.4.1423.
This paper gives three related results: (i) a new, simple, fast, monotonically converging algorithm for deriving the L1-median of a data cloud in Rd, a problem that can be traced to Fermat and has fascinated applied mathematicians for over three centuries; (ii) a new general definition for depth functions, as functions of multivariate medians, so that different definitions of medians will, correspondingly, give rise to different dept functions; and (iii) a simple closed-form formula of the L1-depth function for a given data cloud in Rd.
(i)一种用于推导(R^d)中数据云的(L1)中位数的新的、简单、快速且单调收敛的算法,该问题可追溯到费马,并且在三个多世纪以来一直吸引着应用数学家;(ii)深度函数的一个新的通用定义,作为多元中位数的函数,这样中位数的不同定义将相应地产生不同的深度函数;以及(iii)对于(R^d)中给定数据云的(L1)深度函数的一个简单的封闭形式公式。