Canonical: https://knowlpedia.com/probability/independence-random-variables/

# Independence of random variables

Definition of when random variables have factorizing joint probabilities.

A **family of [random variables](https://knowlpedia.com/probability/random-variable/)** \((X_i)_{i\in I}\) on a [probability space](https://knowlpedia.com/probability/probability-space/) \((\Omega,\mathcal F,\mathbb P)\) is **independent** if for every finite choice of distinct indices \(i_1,\dots,i_k\in I\) and every choice of Borel sets \(A_1,\dots,A_k\subseteq\mathbb R\),
\[
\mathbb P\big(X_{i_1}\in A_1,\dots,X_{i_k}\in A_k\big)=\prod_{j=1}^k \mathbb P\big(X_{i_j}\in A_j\big).
\]

## Equivalent characterizations

Equivalently, the [sigma-algebras](https://knowlpedia.com/measure-theory/sigma-algebra/) \(\sigma(X_i)\) generated by the variables are [independent](https://knowlpedia.com/probability/independence-sigma-algebras/).

## Remarks

This says that all events of the form \(\{X_i\in A\}\) behave like [independent events](https://knowlpedia.com/probability/independence-events/) under [probability](https://knowlpedia.com/probability/event-probability/).

## Examples

- Let \(\Omega=\{0,1\}^2\) with \(\mathbb P\) uniform, and define \(X(\omega_1,\omega_2)=\omega_1\), \(Y(\omega_1,\omega_2)=\omega_2\). Then \(X\) and \(Y\) are independent [random variables](https://knowlpedia.com/probability/random-variable/).
- Let \(\Omega=[0,1]^2\) with the product [Lebesgue measure](https://knowlpedia.com/measure-theory/lebesgue-measure/) (normalized to a [probability measure](https://knowlpedia.com/probability/probability-measure/)), and set \(X(u,v)=u\), \(Y(u,v)=v\). Then \(X\) and \(Y\) are independent and each has the [uniform distribution](https://knowlpedia.com/probability/uniform-distribution/) on \([0,1]\).
