Exercise on measure theory, (verification and suggestion)

122 Views Asked by At

Hi everyone I'd like to know if the following is correct and also I'd appreciate any suggestion to improve the argument. Thanks in advance

For every positive integer $n$, let $f_n:{\bf{R}}\to [0,\infty)$ be a non-negative measurable function such that $$\int_{\bf{R}}f_n\le \frac{1}{4^n}$$ show that for every $\varepsilon>0$ there exists a set $E$ of lebesgue measure $m(E)\le \varepsilon$ such that $f_n$ converges pointwise to zero for all $x\in R\setminus E$.

Pf: Let $A_n=f_n^{-1}\big(\big(1/2^n\varepsilon,\infty\big)\big)$ we will show that $m(A_n)\le \varepsilon/2^n$ for all $n\in {\bf{N}}$. Suppose for sake of contradiction that $m(A_n)>\varepsilon/2^n$ for some $n$. Let $s=1/2^n\varepsilon\cdot 1_{A_n}$ so $$\int_{\bf{R}}f_n\ge \int_{\bf{R}}s=\frac{1}{2^n\varepsilon}m(A_n)>\frac{1}{4^{n}}$$ Contradiction. So for all $n\in {\bf{N}}$ we thus have $m(A_n)\le \varepsilon /2^n$. Let $E =\bigcup_{n\in \bf{N}}A_n$ using subadditivity of $m$ we have $$m(E)\le \sum_{n\in \bf{N}}m(A_n)\le \varepsilon$$

We claim that $E$ is the set we're looking for. Suppose $x\notin E$ this means that $x\notin A_n$ for all $n\in {\bf{N}}$. Thus $f_n(x)\in[0,1/2^n\varepsilon]$ for all $n$, i.e., $f_n(x)\to 0$ as $n\uparrow \infty$ and $x\in {\bf{R}}\setminus E$, i.e., $f_n$ converges pointwise to zero as long as $x\in {\bf{R}}\setminus E$ which concludes the proof.

1

There are 1 best solutions below

4
On BEST ANSWER

Your argument works fine.

Another way to do this is noting that $f=\sum f_n$ defines an $L^1(\mathbb{R})$ function by the monotone convergence theorem, and is thus finite a.e. But then $f_n$ must tend to zero a.e. (at least at every point where the series converges).