#### (PP 4.1) Expectation for discrete random variables

(“Average value”)

Let $\mathscr{X} = X(\Omega)$.

Definition. The expectation of a random variable $X$ with PMF $p$ is

when this sum is ‘well-defined.’ Otherwise, the expactation does not exist.

Well-defined means well-deinfed as a sum of infinite sereis.”

Definition. Let $a_1, a_2, \ldots \in \mathbb{R}$ and let $\sum_{i=1}^\infty a_i = \sum_{i: a_i \ge 0}a_i + \sum_{i: a_i < 0}a_i = b+c$.
$\sum_{i=1}^\infty a_i$ is well-defined if either $b$ or $c$ is finite.

Example. Let $c = \sum_{k = 1}^\infty \frac{1}{k^2} < \infty$

• $\mathbb{E}(X)$ can be inifinite: let $p(k) = \frac{1}{ck^2}$ for $k \in \{1, 2, \ldots \}$
$\mathbb{E}(X) = \sum_{i=1}^\infty \frac{k}{ck^2} = \sum_{k=1}^\infty = \infty$
• $\mathbb{E}(X)$ might not exists: $p(k) = \frac{1}{ck^2}$ for $k \in \mathbb{Z} -\{0\}$ $% $ (undefined)

#### (PP 4.2) Expectation for random variables with densities

Definition. The expectation of a random variable $X$ with density $f$ is

when this integral is ‘well-defined.’ Otherwise, the expactation does not exist.

Example. $X \sim {\rm Uniform}(a,b)$
$\mathbb{E}(X) = \int_a^b\frac{x}{b-a}dx = \frac{1}{b-a} \frac{x^2}{2}\vert_a ^b=\frac{1}{b-a}\frac{1}{2}(b^2-a^2)=\frac{a+b}{2}$

#### (PP 4.3) Expectation rule

Fact. $g(X)$ is a random variable and $g: \mathbb{R} \rightarrow \mathbb{R}$ is measurable.

Theorem. (Expectation rule) If $X$ is a random variable and $g: \mathbb{R} \rightarrow \mathbb{R}$ is measurable, then

1. $\mathbb{E}(g(X)) = \sum_{x \in \mathscr{X}}g(x)p(x)$ if $X$ is discrete with PMF $p$
2. $\mathbb{E}(g(X)) = \int_{-\infty}^\infty g(x)p(x)$ if $X$ is discrete with PMF $p$
(when these quantities are well-defined)

#### (PP 4.4) Properties of expectation

Theorem. Suppose $X, Y$ are random variables such that $E\vert X\vert, E\vert Y\vert <\infty$.

1. $\mathbb{E}(a) = a$ for all $a \in \mathbb{R}$
2. $\mathbb{E}(aX) = a\mathbb{E}(X)$ for all $a \in \mathbb{R}$
3. $\mathbb{E}(X + Y) = \mathbb{E}(X) + \mathbb{E}(Y)$
4. If $X \ge 0$ then $\mathbb{E}(X) \ge 0$
5. If $X \le Y$ then $\mathbb{E}(X) \le \mathbb{E}(Y)$
6. $\mathbb{E}(I_A(X)) = P(X \in A)$

Theorem. If $X_1, X_2, \ldots$ is a sequence of random variables such that $X_i \ge 0$ for all $i$, then

#### (PP 4.5) Mean, variance, and moments

Let $X$ be a random variable.

Definition.

1. The mean $\mu(X = \mathbb{E}(X)$
2. The variance $\sigma^2(X)$, ${\rm Var}(X)$ is $\mathbb{E}((X-\mathbb{E}(X))^2)$
3. The $k$-the moment $m_k(X)$ is $\mathbb{E}(X^k)$
4. The $k$-th central moment is $\mathbb{E}((X-\mathbb{E}(X))^k)$