Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

From Wikipedia, Anderson's theorem is stated as:

Let $K$ be a convex body in n-dimensional Euclidean space $\mathbb{R}^n$ that is symmetric with respect to reflection in the origin, i.e. $K = −K$. Let $f : \mathbb{R}^n \to \mathbb{R}$ be a non-negative, symmetric, globally integrable, unimodal function. Then, for any $0 ≤ c ≤ 1$ and $y ∈ \mathbb{R}^n$, $$ \int_{K} f(x + c y) \, \mathrm{d} x \geq \int_{K} f(x + y) \, \mathrm{d} x. $$

Then Wikipedia also gives its application to probability theory:

Given a probability space $(Ω, Σ, P)$, suppose that $X : Ω → \mathbb{R}^n$ is a random variable with probability density function $f : \mathbb{R}^n → [0, +∞)$ and that $Y : Ω → \mathbb{R}^n$ is an independent random variable. The probability density functions of many well-known probability distributions are p-concave for some p, and hence unimodal. If they are also symmetric, then Anderson's theorem applies, in which case

$$ P ( X \in K ) \geq P ( X + Y \in K ) $$ for any origin-symmetric convex body $K ⊆ \mathbb{R}^n$.

I was wondering

  1. How does Anderson's theorem lead to the inequality for probability, given $Y$ is not a constant but a random variable? The density function of $X+Y$ is the convolution of the densities of $X$ and of $Y$, which has a different form from the density of $X$. Also I am not sure if $Y$ may allow to not have a density.
  2. What are the requirements on the distributions of $X$ and $Y$? The quote is a bit vague to me? Does it say that the distributions of $X$ and of $Y$ should be both unimodal and symmetric?
  3. How is $p$-concavity defined for a distribution and for a general function? I found different definitions on the internet and am not sure which one is assumed here.

I also appreciate if there can be some pointers to references regarding Anderson's theorem and its proof.

Thanks and regards!

share|improve this question
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.