Begrenzende Summe der iid Gamma-Variationen

Sei $X_1,X_2,\ldots$ eine Folge von unabhängig und identisch verteilten Zufallsvariablen mit der Wahrscheinlichkeitsdichtefunktion;

f (x) = {\begin{cases} \frac{1}{2} x^{2} e^{- x} & if x > 0; \\ 0 & otherwise . \end{cases}

$f(x) = \left\{ \begin{array}{ll} \frac{1}{2}x^2 e^{-x} & \mbox{if $x>0$};\\ 0 & \mbox{otherwise}.\end{array} \right.$ Zeigen Sie, dass

lim_{n \to \infty} P [X_{1} + X_{2} + \dots + X_{n} \geq 3 (n - \sqrt{n})] \geq \frac{1}{2}

$\lim_{n\to \infty} P[X_1+X_2+\ldots+X_n\ge 3(n-\sqrt{n})] \ge \frac{1}{2}$

Was ich versucht habe

Auf den ersten Blick dachte ich, es sollte Chebyshevs Ungleichung verwenden, da die Frage eine Untergrenze . Ich dachte jedoch über das Grenzwertzeichen nach, das eindeutig anzeigt, dass das Problem in irgendeiner Weise mit dem zentralen Grenzwertsatz (CLT) zusammenhängen kann. $X_1+X_2+\ldots +X_n$

Sei $S_n=X_1+X_2+\ldots +X_n$

E (S_{n}) = \sum_{i = 0}^{n} E (X_{i}) = 3 n (since E (X_{i}) = 3) V (S_{n}) = \sum_{i = 0}^{n} V (X_{i}) = 3 n (since V (X_{i}) = 3 and X_{i} are i.i.d)

$E(S_n)=\sum_{i=0}^{n} E(X_i)=3n \ (\text{since } E(X_i)=3) \\ V(S_n)=\sum_{i=0}^{n} V(X_i)=3n \ (\text{since } V(X_i)=3 \text{ and } X_i \text{ are i.i.d})$

Verwenden Sie jetzt CLT für großes , Oder $n$ $X_1+X_2+........+X_n \sim N(3n,3n)$

z = \frac{S_{n} - 3 n}{\sqrt{3 n}} \sim N (0, 1) as n \to \infty

$z=\frac{S_n-3n}{\sqrt{3n}} \sim N(0,1) \text{ as } n\to \infty$

Nun ist

lim_{n \to \infty} P [X_{1} + X_{2} + . . . . . . . . + X_{n} \geq 3 (n - \sqrt{n})] = lim_{n \to \infty} P (S_{n} - 3 n \geq - 3 \sqrt{n}) = lim_{n \to \infty} P (\frac{S_{n} - 3 n}{\sqrt{3 n}} \geq - \sqrt{3}) = P (z \geq - \sqrt{3}) = P (- \sqrt{3} \leq z < 0) + P (z \geq 0) = P (- \sqrt{3} \leq z < 0) + \frac{1}{2} \dots (1)

$\lim_\limits{n\to \infty} P[X_1+X_2+........+X_n\ge 3(n-\sqrt{n})] = \lim_\limits{n\to \infty}P(S_n-3n \ge -3\sqrt{n}) = \lim_\limits{n\to \infty} P\left(\frac{S_n-3n}{\sqrt{3n}} \ge -\sqrt{3}\right) =P(z\ge -\sqrt{3}) =P\left(-\sqrt{3}\le z<0\right)+P(z\ge 0 ) =P\left(-\sqrt{3}\le z<0\right)+\frac{1}{2}\cdots(1)$

Da , also aus, $P(-\sqrt{3}\le z<0) \ge 0$ $(1)$

lim_{n \to \infty} P [X_{1} + X_{2} + . . . . . . . . + X_{n} \geq 3 (n - \sqrt{n})] \geq \frac{1}{2}

$\lim_\limits{n\to \infty} P[X_1+X_2+........+X_n\ge 3(n-\sqrt{n})]\ge \frac{1}{2}$

Hab ich recht?

probability self-study

CLT scheint einen vernünftigen Ansatz , aber "

"macht keinen Sinn ..

lim_{n \to \infty} P [X_{1} + X_{2} + . . . . . . . . + X_{n} \geq 3 (n - \sqrt{n})] = P (S_{n} - 3 n \geq - 3 \sqrt{n})

$\lim_\limits{n\to \infty} P[X_1+X_2+........+X_n\ge 3(n-\sqrt{n})] =P(S_n-3n \ge -3\sqrt{n})$

— P. Windridge

Ich denke, es sollte

lim_{n \to \infty} P [X_{1} + X_{2} + . . . . . . . . + X_{n} \geq 3 (n - \sqrt{n})] = lim_{n \to \infty} P (S_{n} - 3 n \geq - 3 \sqrt{n}) = lim_{n \to \infty} P (\frac{S_{n} - 3 n}{\sqrt{3 n}} \geq - \sqrt{3}) = P (z \geq - \sqrt{3})

X_{i} \sim Γ (3, 1)

$X_i \sim\Gamma(3,1)$

X_{1} + X_{2} + \cdot + X_{n} \sim Γ (3 n, 1)

$X_1+X_2+\cdot+X_n \sim \Gamma(3n,1)$

n

$n$

Γ (3 n, 1)

$\Gamma(3n,1)$

3 n - \frac{1}{3}

$3n-\frac 13$

3 n

$3n$

3 (n - \sqrt{n}) < 3 n - \frac{1}{3}

$3(n-\sqrt{n}) < 3n-\frac 13$

3 (n - \sqrt{n})

$3(n-\sqrt{n})$

$n$

$X_i$ $\mu_i$ $\sigma_i^2.$ $Y_n$ $n$

Y_{n} = \sum_{i = 1}^{n} X_{i} .

$Y_n = \sum_{i=1}^n X_i.$

$Y_n$

m_{n} = \sum_{i = 1}^{n} μ_{n}

$m_n = \sum_{i=1}^n \mu_n$

und seine Varianz ist

s_{n}^{2} = Var (Y_{n}) = \sum_{i = 1}^{n} Var (X_{i}) + 2 \sum_{j > i} Cov (X_{i}, X_{j}) = \sum_{i = 1}^{n} σ_{i}^{2} .

$s_n^2 = \operatorname{Var}(Y_n) = \sum_{i=1}^n \operatorname{Var}(X_i) + 2\sum_{j > i}\operatorname{Cov}(X_i,X_j) = \sum_{i=1}^n \sigma_i^2.$

Suppose $s_n^2$ grows at most linearly with $n$ : that is, there exists a number $\lambda\gt 0$ such that for all sufficiently large $n,$ $s_n^2 \le \lambda^2 n.$ Let $k\gt 0$ (yet to be determined), observe that

m - k \sqrt{n} \leq m - \frac{k}{λ} s_{n},

$m - k \sqrt{n} \le m - \frac{k}{\lambda}s_n,$

and apply Chebyshev's Inequality to $Y_n$ to obtain

\begin{aligned} Pr (Y_{n} \geq m_{n} - k \sqrt{n}) & \geq Pr (Y_{n} \geq m_{n} - \frac{k}{λ} s_{n}) \\ \geq Pr (| Y_{n} - m_{n} | \leq \frac{k}{λ} s_{n}) \\ \geq 1 - \frac{λ^{2}}{k^{2}} . \end{aligned}

$\eqalign{ \Pr(Y_n \ge m_n - k\sqrt{n}) &\ge \Pr(Y_n \ge m_n - \frac{k}{\lambda}s_n) \\ &\ge \Pr(|Y_n - m_n| \le \frac{k}{\lambda} s_n) \\ &\ge 1 - \frac{\lambda^2}{k^2}. }$

The first two inequalities are basic: they follow because each successive event is a subset of the preceding one.

In the case at hand, where $X_i$ are independent (and therefore uncorrelated) with means $\mu_i=3$ and variances $\sigma_i^2=3,$ we have $m_n=3n$ and

s_{n} = \sqrt{3} \sqrt{n},

$s_n = \sqrt{3}\sqrt{n},$

whence we may take $\lambda$ as small as $\sqrt{3}.$ The event in the question $3(n-\sqrt{n}) = \mu_n - 3\sqrt{n}$ corresponds to $k=3,$ where

Pr (Y_{n} \geq 3 n - 3 \sqrt{n}) \geq 1 - \frac{{\sqrt{3}}^{2}}{3^{2}} = \frac{2}{3} > \frac{1}{2},

$\Pr(Y_n \ge 3n - 3\sqrt{n}) \ge 1 - \frac{\sqrt{3}^{\ 2}}{3^2} = \frac{2}{3}\gt \frac{1}{2},$

QED.

— whuber
quelle