Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
joint normal distribution∗ gel† 2013-03-21 19:24:27 A finite set of random variables X1 , . . . , Xn are said to have a joint normal distribution or multivariate normal distribution if all real linear combinations λ1 X1 + λ2 X2 + · · · + λn Xn are normal. This implies, in particular, that the individual random variables Xi are each normally distributed. However, the converse is not not true and sets of normally distributed random variables need not, in general, be jointly normal. If X = (X1 , X2 , . . . , Xn ) is joint normal, then its probability distribution is uniquely determined by the means µ ∈ Rn and the n × n positive semidefinite covariance matrix Σ, µi = E[Xi ], Σij = Cov(Xi , Xj ) = E[Xi Xj ] − E[Xi ]E[Xj ]. Then, the joint normal distribution is commonly denoted as N(µ, Σ). Conversely, this distribution exists for any such µ and Σ. Figure 1: Density of joint normal variables X, Y with Var(X) = 2, Var(Y ) = 1 and Cov(X, Y ) = −1. The joint normal distribution has the following properties: 1. If X has the N(µ, Σ) distribution for nonsigular Σ then it has the multidimensional Gaussian probability density function 1 1 fX (x) = p exp − (x − µ)T Σ−1 (x − µ) . 2 (2π)n det (Σ) 2. If X has the N(µ, Σ) distribution and λ ∈ Rn then λ · X = λ1 X1 + · · · + λn Xn ∼ N(λ · µ, λT Σλ). ∗ hJointNormalDistributioni created: h2013-03-21i by: hgeli version: h37204i Privacy setting: h1i hDefinitioni h62H05i h60E05i † This text is available under the Creative Commons Attribution/Share-Alike License 3.0. You can reuse this document or portions thereof only if you do so under terms that are compatible with the CC-BY-SA license. 1 3. Sets of linear combinations of joint normals are themselves joint normal. In particular, if X ∼ N(µ, Σ) and A is an m × n matrix, then AX has the joint normal distribution N(Aµ, AΣAT ). 4. The characteristic function is given by 1 T ϕX (a) ≡ E [exp(ia · X)] = exp ia · µ − a Σa , 2 for X ∼ N(µ, Σ) and any a ∈ Cn . 5. A pair X, Y of jointly normal random variables are independent if and only if they have zero covariance. 6. Let X be a random vector whose distribution is jointly normal. Suppose the coordinates of X are partitioned into two groups, forming random vectors X1 and X2 , then the conditional distribution of X1 given X2 = c is jointly normal. 2