Definition:Weighted Mean

From ProofWiki
Jump to navigation Jump to search

Definition

Let $S = \sequence {x_1, x_2, \ldots, x_n}$ be a sequence of real numbers.

Let $W$ be a weight function to be applied to the terms of $S$.


The weighted mean of $S$ with respect to $W$ is defined as:

$\bar x := \dfrac {\ds \sum_{i \mathop = 1}^n \map W {x_i} x_i} {\ds \sum_{i \mathop = 1}^n \map W {x_i} }$

This means that elements of $S$ with a larger weight contribute more to the weighted mean than those with a smaller weight.


If we write:

$\forall i: 1 \le i \le n: w_i = \map W {x_i}$

we can write this weighted mean as:

$\bar x := \dfrac {w_1 x_1 + w_2 x_2 + \cdots + w_n x_n} {w_1 + w_2 + \cdots + w_n}$


From the definition of the weight function, none of the weights can be negative.

While some of the weights may be zero, not all of them can, otherwise we would be dividing by zero.


Normalized Weighted Mean

Let the weights be normalized.

Then the weighted mean of $S$ with respect to $W$ can be expressed in the form:

$\ds \bar x := \sum_{i \mathop = 1}^n \map W {x_i} x_i$

as by definition of normalized weight function all the weights add up to $1$.

This weighted mean is known as a normalized weighted mean.


Also see

  • Results about weighted means can be found here.


Sources

Retrieved from "https://proofwiki.org/w/index.php?title=Definition:Weighted_Mean&oldid=699401"