Sample variance and standard deviation

L. El Ghaoui

Sample variance and standard deviation

The sample variance of given numbers $x_1, \cdots, x_n$ , is defined as

$\sigma^2 := \frac{1}{n} ((x_1 -\hat{x})^2 + \cdots + (x_n - \hat{x})^2),$

where $\hat{x}$ is the sample average of $x_1, \cdots, x_n$ . The sample variance is a measure of the deviations of the numbers $x_i$ with respect to the average value $\hat{x}$ .

The sample standard deviation is the square root of the sample variance, $\sigma^2$ . It can be expressed in terms of the Euclidean norm of the vector $x = (x_1, \cdots, x_n)$ , as

$\sigma = \frac{1}{\sqrt{n}}||x-\hat{x}{\bf 1}||_2,$

where $||\cdot||_2$ denotes the Euclidean norm.

More generally, for any vector $p \in \mathbb{R}^n$ , with $p_i \ge 0$ for every $i$ , and $p_1 + \cdots + p_n = 1$ , we can define the corresponding weighted variance as

$\sum\limits_{i=1}^n p_i(x_i - \hat{x})^2.$

The interpretation of $p$ is in terms of a discrete probability distribution of a random variable $X$ , which takes the value $x_i$ with probability $p_i$ , $i = 1, \cdots, n$ . The weighted variance is then simply the expected value of the squared deviation of $X$ from its mean ${\bf E}(X)$ , under the probability distribution $p$ .

See also: sample and weighted average.

Sample variance and standard deviation

License

Share This Book