Extra Problems for Chapter 11

T.M. Cover and J.A. Thomas

  1. Mean squared error.
    Let tex2html_wrap_inline419 satisfy tex2html_wrap_inline421 . Consider linear predictors for tex2html_wrap_inline423 , i.e.

    displaymath413

    Assume n>p. Find

    displaymath414

    where the minimum is over all linear predictors b and the maximum is over all densities f satisfying tex2html_wrap_inline431 .

  2. Maximum entropy characteristic functions.
    We ask for the maximum entropy density tex2html_wrap_inline433 satisfying a constraint on the characteristic function tex2html_wrap_inline435 . The answers need be given only in parametric form.
    1. Find the maximum entropy f satisfying tex2html_wrap_inline439 , at a specified point tex2html_wrap_inline441 .
    2. Find the maximum entropy f satisfying tex2html_wrap_inline445 .
    3. Find the maximum entropy density tex2html_wrap_inline433 having a given value of the characteristic function tex2html_wrap_inline449 at a specified point tex2html_wrap_inline441 .
    4. What problem is encountered if tex2html_wrap_inline453 ?
  3. Maximum entropy processes.
    1. Find the maximum entropy rate binary stochastic process tex2html_wrap_inline455 , satisfying tex2html_wrap_inline457 for all i.
    2. What is the resulting entropy rate?
  4. Maximum entropy of sums Let tex2html_wrap_inline461 Find the maximum entropy density for Y under the constraint tex2html_wrap_inline465 , tex2html_wrap_inline467 ,
    1. if tex2html_wrap_inline469 and tex2html_wrap_inline471 are independent.
    2. if tex2html_wrap_inline469 and tex2html_wrap_inline471 are allowed to be dependent.
    3. Prove part (a).
  5. Maximum entropy Markov chain.

    Let tex2html_wrap_inline477 be a stationary Markov chain with tex2html_wrap_inline479 . Let tex2html_wrap_inline481 for all n.

    1. What is the maximum entropy rate process satisfying this constraint?
    2. What if tex2html_wrap_inline485 , for all n for some given value of tex2html_wrap_inline489 , tex2html_wrap_inline491 ?
  6. An entropy bound on prediction error. Let tex2html_wrap_inline493 be an arbitrary real valued stochastic process. Let tex2html_wrap_inline495 . Thus the conditional mean tex2html_wrap_inline497 is a random variable depending on the n-past tex2html_wrap_inline501 . Here tex2html_wrap_inline497 is the minimum mean squared error prediction of tex2html_wrap_inline505 given the past.
    1. Find a lower bound on the conditional variance tex2html_wrap_inline507 in terms of the conditional differential entropy tex2html_wrap_inline509 .
    2. Is equality achieved when tex2html_wrap_inline493 is a Gaussian stochastic process?
  7. Maximum entropy rate. What is the maximum entropy rate stochastic process tex2html_wrap_inline477 over the symbol set tex2html_wrap_inline515 for which the probability that 00 occurs in a sequence is zero?
  8. Maximum entropy.
    1. What is the parametric form maximum entropy density f(x) satisfying the two conditions

      displaymath267

    2. What is the maximum entropy density satisfying the condition

      displaymath273

    3. Which entropy is higher?
  9. Maximum entropy. Find the parametric form of the maximum entropy density f satisfying the Laplace transform condition

    displaymath415

    and give the constraints on the parameter.



Latex file

Postscript file


 

Joy Thomas
Sat Aug 15 07:52:28 EDT 1998