Usbrth

Multi tool use

XeLaTeX and pdfLaTeX ignore hyphenation

I see my dog run

Why is an old chain unsafe?

New order #4: World

Why CLRS example on residual networks does not follows its formula?

How can the DM most effectively choose 1 out of an odd number of players to be targeted by an attack or effect?

Circuitry of TV splitters

Why do we use polarized capacitor?

How to type dʒ symbol (IPA) on Mac?

Can an x86 CPU running in real mode be considered to be basically an 8086 CPU?

declaring a variable twice in IIFE

Copycat chess is back

Schwarzchild Radius of the Universe

How do you conduct xenoanthropology after first contact?

Why is the design of haulage companies so “special”?

What are these boxed doors outside store fronts in New York?

I’m planning on buying a laser printer but concerned about the life cycle of toner in the machine

Is it possible to do 50 km distance without any previous training?

I probably found a bug with the sudo apt install function

Shell script can be run only with sh command

Simulate Bitwise Cyclic Tag

Calculus Optimization - Point on graph closest to given point

What would happen to a modern skyscraper if it rains micro blackholes?

Is it possible to make sharp wind that can cut stuff from afar?

Moving Supremum inside conditional expectation

Conditions under which the Limit for “Measure $to 0$” is $0$ [more difficult]Conditional expectation is square-integrableProperty of conditional expectationDifferentiating inside a conditional expectationEquality of sets supremum and infimumConvergence of conditional meansExchanging supremum and conditional expectationProve or disprove the following inequality?Conditional expectation as a random variable newSupremum of integrable random variables and pointwise convergence

0

$begingroup$

Let $Z_l_l in Lambda,X$ be square-integrable random-variables, and $f:mathbbR^2rightarrow mathbbR$ be a continuous function, such that

• $
  sup_l in LambdamathbbE[f(Z_l,X)]<infty
  .
  $


• $operatornameesssup_l in LambdaZ_l<infty$

Under what conditions, is it true that
$$
sup_l in LambdamathbbEleft[
f(Z_l,X)
right]
=
mathbbEleft[
f(operatornameesssup_l in LambdaZ_l,X)
right]
$$

probability probability-theory measure-theory calculus-of-variations supremum-and-infimum

share|cite|improve this question

asked Apr 1 at 15:56

AIM_BLBAIM_BLB

2,5392820

$endgroup$

• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  I think you are talking about the normal expected value, since conditional expectation is a random variable itself. At first glance, you need monotony in $f$ for $Z_l$, i.e. $Z_ileq Z_k$ implies $mathbbE[f(Z_i,X)] leq mathbbE[f(Z_k,X)]$. The bounds guarantee you that such a $sup$ exists almost surely (inside as well as outside of the expected value), so you are fine taking a surpremum over a bounded sequence. This is just a quick guess, maybe someone else is able to provide you with a more detailed answer with more relaxed conditions.
  $endgroup$
  – F. Conrad
  Apr 1 at 22:03
  
  

  
  
  
  

add a comment | 

0

$begingroup$

Let $Z_l_l in Lambda,X$ be square-integrable random-variables, and $f:mathbbR^2rightarrow mathbbR$ be a continuous function, such that

• $
  sup_l in LambdamathbbE[f(Z_l,X)]<infty
  .
  $


• $operatornameesssup_l in LambdaZ_l<infty$

Under what conditions, is it true that
$$
sup_l in LambdamathbbEleft[
f(Z_l,X)
right]
=
mathbbEleft[
f(operatornameesssup_l in LambdaZ_l,X)
right]
$$

probability probability-theory measure-theory calculus-of-variations supremum-and-infimum

share|cite|improve this question

asked Apr 1 at 15:56

AIM_BLBAIM_BLB

2,5392820

$endgroup$

• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  I think you are talking about the normal expected value, since conditional expectation is a random variable itself. At first glance, you need monotony in $f$ for $Z_l$, i.e. $Z_ileq Z_k$ implies $mathbbE[f(Z_i,X)] leq mathbbE[f(Z_k,X)]$. The bounds guarantee you that such a $sup$ exists almost surely (inside as well as outside of the expected value), so you are fine taking a surpremum over a bounded sequence. This is just a quick guess, maybe someone else is able to provide you with a more detailed answer with more relaxed conditions.
  $endgroup$
  – F. Conrad
  Apr 1 at 22:03
  
  

  
  
  
  

add a comment | 

$begingroup$

Let $Z_l_l in Lambda,X$ be square-integrable random-variables, and $f:mathbbR^2rightarrow mathbbR$ be a continuous function, such that

• $
  sup_l in LambdamathbbE[f(Z_l,X)]<infty
  .
  $


• $operatornameesssup_l in LambdaZ_l<infty$

Under what conditions, is it true that
$$
sup_l in LambdamathbbEleft[
f(Z_l,X)
right]
=
mathbbEleft[
f(operatornameesssup_l in LambdaZ_l,X)
right]
$$

probability probability-theory measure-theory calculus-of-variations supremum-and-infimum

share|cite|improve this question

asked Apr 1 at 15:56

AIM_BLBAIM_BLB

2,5392820

$endgroup$

share|cite|improve this question

asked Apr 1 at 15:56

AIM_BLBAIM_BLB

2,5392820

share|cite|improve this question

asked Apr 1 at 15:56

AIM_BLBAIM_BLB

2,5392820

asked Apr 1 at 15:56

AIM_BLBAIM_BLB

2,5392820

asked Apr 1 at 15:56

AIM_BLBAIM_BLB

2,5392820