MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Gauss's inequality is for unimodal distributions, concerning distance from the mode.

A similar result is Vysochanskiï–Petunin inequality, which is for the distance from the mean rather than the mode. Chebyshev's inequality generalizes Vysochanskiï–Petunin inequality by concerning distance from the mean without requiring unimodality.

I wonder if there is generalization of Gauss's inequality for distributions not necessarily unimodal?


share|cite|improve this question
up vote 2 down vote accepted

Chebyshev's inequality is crude: The probability of being more than $k$ standard deviations away from the mean is at most $1/k^2$. A similar coarse estimate applies to the distance from the mode or any other point. Let $r_x^2$ be the average of the square of the distance from $x$, which can be computed as $(x-\mu)^2 + \sigma^2$. Then the probability of being more than $kr_x$ away from $x$ is at most $1/k^2$. This is true for all $x$, including a mode. The same works if we let $x$ denote a set, including a set of modes, although computing $r_x^2$ becomes harder.

share|cite|improve this answer
Thanks! How is $r_x^2$ defined when x is a set? – Tim May 25 '12 at 1:01
The distance between the point $y$ and the set $X$ is $d(y,X) = \inf_{x \in X} d(y,x)$. Then $r_X^2$ can be defined as the average of the square of the distance to $X$. This is no longer a quadratic polynomial so it can't be computed as easily as $(x-\mu)^2 + \sigma^2$, but the inequality still holds. – Douglas Zare May 25 '12 at 1:30

Your Answer


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

Not the answer you're looking for? Browse other questions tagged or ask your own question.