Missing $g$-mass: Investigating the Missing Parts of Distributions

TL;DR

This paper introduces the concept of missing g-mass to analyze unobserved parts of large distributions, providing new estimation techniques and concentration bounds for various functions of the missing distribution.

Contribution

It defines missing g-mass, studies minimax estimation for order-alpha missing mass, and develops new concentration bounds including strongly sub-Gamma and filtered sub-Gaussian types.

Findings

Exact minimax convergence rates for order-alpha missing mass.

Sub-Gaussian tail bounds with near-optimal variance factors.

Introduction of strongly sub-Gamma and filtered sub-Gaussian concentration notions.

Abstract

Estimating the underlying distribution from \textit{iid} samples is a classical and important problem in statistics. When the alphabet size is large compared to number of samples, a portion of the distribution is highly likely to be unobserved or sparsely observed. The missing mass, defined as the sum of probabilities $\text{Pr}(x)$ over the missing letters $x$, and the Good-Turing estimator for missing mass have been important tools in large-alphabet distribution estimation. In this article, given a positive function $g$ from $[0,1]$ to the reals, the missing $g$-mass, defined as the sum of $g(\text{Pr}(x))$ over the missing letters $x$, is introduced and studied. The missing $g$-mass can be used to investigate the structure of the missing part of the distribution. Specific applications for special cases such as order-$\alpha$ missing mass ($g(p)=p^{\alpha}$) and the missing Shannon entropy ($g(p)=-p\log p$) include estimating distance from uniformity of the missing distribution and its partial estimation. Minimax estimation is studied for order-$\alpha$ missing mass for integer values of $\alpha$ and exact minimax convergence rates are obtained. Concentration is studied for a class of functions $g$ and specific results are derived for order-$\alpha$ missing mass and missing Shannon entropy. Sub-Gaussian tail bounds with near-optimal worst-case variance factors are derived. Two new notions of concentration, named strongly sub-Gamma and filtered sub-Gaussian concentration, are introduced and shown to result in right tail bounds that are better than those obtained from sub-Gaussian concentration.