- Markov’s Inequality; Chebychev’s Inequality; Chernoff’s Bound.
- Bounds for the Poisson Distribution.
Throughout this section we assume that is a real valued random variable. The mean, variance and moment generating function of
are, respectively,
,
and
, when these exist.
Ex 1. [Markov’s Inequality] For positive random variable
Markov’s inequality can be thought of as an upper-bound on the probability or as a lower-bound on the expectation:
Ex 2. Let be the mode of
, i.e.
then
(Similar logic applies to other quartiles.)
Ex 3. [Chebychev’s Inequality] For with mean
and variance
Ex 4. For a positive increasing function ,
Ex 5. [Chernoff’s Bound]
Let’s get some handy exponential and normal tail bounds for the Poisson distribution.
Ex 6. [Some Poisson Bounds] For , a Poisson random variable with parameter
,
and
Ex 7. [Continued]
Ex 8.[Continued] For ,
Answers.
Ans 1. now take expectations.
Ans 2. Applying Markov’s inequality
Ans 3. Consider , square both sides and apply Markov [1].
Ans 4. Consider , apply
to both sides and then [1].
Ans 5. Take apply [4], then minimize over theta.
Ans 6. First equality is straightforward. Second, implies
which after substitution gives the bound.
Ans 7. First bound: The Chernoff bound [5] gives
By assumption , substituting this into the logorithm above gives the result. Second, by the same Chernoff bound and use that
.
Ans 8. Similar to [7] with ; however now apply the bound
.