Measure Theory/Markov and Hardy-Littlewood

From testwiki
Jump to navigation Jump to search

Markov's Inequality

We've already had enough preamble for this theorem. So let's just jump in and state, then prove, the following theorem.

Theorem: Let f: be an integrable function and c+.

Define Ec={x:c|f(x)|}. (Think of this as the set of points at which f is large.)

Then

λ(Ec)1cf

Template:Robelbox

Prove Markov's inequality by construing λ(Ec) as the integral of 𝟏Ec, and then apply the ML bound.

(Would that it were so simple to prove the Hardy-Littlewood inequality.)

Template:Robelbox/close

Hardy-Littlewood

Let Ec={x:c<f*(x)} and we want to show that

λ(Ec)<3cf

To start on the proof, we first observe that if xEc is arbitrary then there is some t+ such that 12t(xt,x+t)|f|>c.

This allows us to define, for each x, the corresponding open interval (xt,x+t) with t as above. This looks like, perhaps, a cover of Ec by open intervals! That sounds provocative and familiar.

However, it would be senseless to cover for all of Ec, since there is no guarantee that Ec is compact. Compactness is precisely what would make an open interval cover useful.

Whence we let FEc be any compact subset. The idea behind what we will do for the remainder of this proof, is to show that λ(F)3c|f|. It will then follow that λ(Ec)3c|f|, which you will prove in an exercise below.

Now for each xF we define the corresponding open interval Ix=(xt,x+t) such that 12tIxf*>c. By the compactness of F, there is a finite subcover, which we will choose to call J1,,Jn.

We would like to reason as follows (although, of course, I would only phrase it this way if there is an obstacle coming):

λ(F)i=1nλ(Ji)=i=1n2ti where ti is the width of interval Ji<i=1n1cJi|f|1c|f|

If this argument were correct then we could get a smaller bound on λ(F), using 1c|f| rather than 3c|f|.

However, the last inequality is not justified because the intervals J1,,Jn may overlap.

Template:Robelbox

In this exercise you will justify each of the steps in

λ(F)i=1nλ(Ji)=i=1n2ti where ti is the width of interval Ji<i=1n1cJi|f|1c|f|

except for the last one, which we have observed is actually invalid.

1. Explain why λ(F)i=1nλ(Ji).

2. Explain why λ(Ji)=2ti.

3. Explain why 2ti<1cJi|f|. Hint: Recall the defining property of Ji.

Template:Robelbox/close

The Vitali Covering Lemma

In order to overcome the obstacle above, that J1,,Jn may fail to be disjoint, we will try to relate this collection to some other collection which is disjoint.

One strategy would be to simply merge overlapping intervals. For instance, if J1,J2 overlap each other, we could replace the pair with J=J1J2. Repeating the procedure finitely many times would produce a new family which is now composed of disjoint intervals.

However, notice that if we did so, the we would no longer be able to say λ(Ji)=2ti.

So we have two competing needs: The need for the intervals to be disjoint but also the need for the intervals to maintain their size.

The easiest resolution is to take the interval in J1,,Jn with the greatest length (ties may be broken arbitrarily), assume without loss of generality that this is J1. If this intersects any other interval then we simply remove those intersecting intervals.

Now define 3*J1 to be the interval with the same center as J1, but with three times its length.

Template:Robelbox

Let I=(a,b), J=(c,d) be two open bounded intervals. Assume that (I)<(J) and that they intersect, IJ.

Prove that I3*J.

Template:Robelbox/close

Template:Robelbox

Let I1,,Im be any family of open bounded intervals. Show that there exists a sub-family Ik1,,Ikn with the properties

  • the family is pairwise disjoint, and
  • i=1mIii=1n3*Iki

This sub-family is called the Vitali-covering for the family I1,,Im

Template:Robelbox/close

The End

Template:Robelbox

Using the Vitali-covering for the family J1,,Jn as in the initial "false proof", correct the false proof to obtain a correct proof that

λ(F)3c|f|

and then conclude the theorem.

Template:Robelbox/close










Template:CourseCat