The Setting
The construction of the signed walk runs a stationary Gaussian autoregressive chain , taken coordinatewise in coordinates, over the steps , driven by a standard Gaussian sample in , and needs a lower bound on the probability that its path satisfies many constraints at once. Each constraint asks that a linear image of the sample lie in a closed convex set symmetric about the origin, and the bound wanted is the product of the probabilities of the single constraints. For a centered Gaussian measure this is Royen’s Gaussian correlation inequality: symmetric convex events are positively correlated.
The library states the inequality in finite dimensions in two forms: for quadratic boxes, the sets cut out by one quadratic inequality in each coordinate, under a centered Gaussian measure with an arbitrary covariance matrix, and for closed symmetric convex sets under linear images of a standard Gaussian measure.
Gaussian Measures, Boxes and Constraint Sets
Let be a finite index set. For a positive semidefinite matrix , is the centered Gaussian measure on with covariance matrix . For , the quadratic box is the subset of (1)✦ and for a set of coordinates, is the set of the with for every .
The standard Gaussian measure of a finite-dimensional real inner product space is the probability measure on under which the coordinates in an orthonormal basis are independent standard normal variables. For a linear map into a real normed space , the image measure is the measure on with . A subset of a real vector space is centrally symmetric when for every .
For a family of continuous linear maps between real normed spaces and and real numbers , indexed by a set , the norm constraint set is the subset of (2)✦
The Claim
Theorem 1. Let be a finite index set, a positive semidefinite matrix, and . Then (3)✦ Let be a finite-dimensional real inner product space. If and are closed, convex, centrally symmetric subsets of , then (4)✦ Let be a linear map into a real normed space . If and are closed, convex, centrally symmetric subsets of , then (5)✦ and if is a finite family of closed, convex, centrally symmetric subsets of , then (6)✦
In the paper, the result reads: if is a centered Gaussian process on with continuous paths and each is a finite intersection of events , with linear, closed, convex and centrally symmetric, and a closed interval, then .
In discrete time, a path event of the chain is a finite intersection of norm constraint sets (2) in its standard Gaussian sample, to which (6) applies.
What Is Checked
The inequalities (3) to (6) are stated over the centered Gaussian measure of a positive semidefinite matrix and the quadratic boxes (1), and over the standard Gaussian measure of a finite-dimensional real inner product space, its images under linear maps, and closed, convex, centrally symmetric sets.