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.

For upper bound of probability $P\{X \ge t\}$, we have Markov's inequality, Chernoff's bound, and moment bound.

But how can we deal with lower bound? Is there any similar inequalities for lower bound analysis?

share|improve this question
    
Chebyshev’s inequality? –  user2468 Apr 1 '12 at 23:44
2  
A discrete random variable with probability masses $\frac{1}{2}$ at $\mu \pm \sigma$ has mean $\mu$ and variance $\sigma^2$, and so $P\{X > t\} = 0$ for all $t > \mu + \sigma$ for this variable. Thus it is going to be difficult to get a lower bound better than $0$ that applies to random variables in general, though with additional conditions one might get a lower bound. –  Dilip Sarwate Apr 2 '12 at 0:18
    
@J.D. Chebyshev's inequality is still an upper bound on probability (for a non-negative variable). It is where Markov's inequality is derived from, so it's likely that the poster already knew of this bound regardless. I suppose one can expand the absolute value usually used in it into two events, then use complements and multiply by -1 to get a statement of lower bounded probability, but it won't be in terms of a useful bound as it will always involve the other probability of the other "half-event" that you split out of the absolute value. –  prpl.mnky.dshwshr Apr 2 '12 at 0:39

1 Answer 1

up vote 4 down vote accepted

Since the indicator function $I_{x \ge t} \ge 1 - (x-a)^2/(t-a)^2$ for $a > t$, $$P(X \ge t) \ge 1 - \frac{\sigma^2+(a-\mu)^2}{(t-a)^2}$$ If $\mu > t$, the optimal $a = \mu + \sigma^2/(\mu-t)$, obtaining $$P(X \ge t) \ge \frac{(\mu-t)^2}{(\mu-t)^2+\sigma^2} \ \text{for}\ t < \mu$$

share|improve this answer

Your Answer

 
discard

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.