Chi-Squared distribution¶

Chi squared is related to the distribution of the sample variance, and can be used to estimate the true variance of the distribution.

\[ S \sim X^2_v \]

Moments¶

Mean¶

\[E[S] = v\]

Variiance¶

\[Var(S) = 2v\]

Connection to Gaussian¶

A Chi-Square with \(v\) degrees of freedom can be defined as a sum of \(v\) standard normals.

\[\begin{split} S \sim X^2_v \\ S = \sum_{i=1}^{v} Z_i^2 \\ Z_i \sim N(0,1)\end{split}\]

Connection to Gamma¶

\[X^2_v \triangleq Ga(\frac{v}{2}, \frac{1}{2})\]