Content deleted Content added
m →Decomposition: The cholesky section implied that all positive definite matrices were able to be decomposed into LL^* and did not specify that only Hermitian matrices were able to be decomposed this way |
add comment about cone |
||
Line 17:
Positive-definite and positive-semidefinite real matrices are at the basis of [[convex optimization]], since, given a [[function of several real variables]] that is twice [[differentiable function|differentiable]], then if its [[Hessian matrix]] (matrix of its second partial derivatives) is positive-definite at a point {{mvar|p}}, then the function is [[convex function|convex]] near {{mvar|p}}, and, conversely, if the function is convex near {{mvar|p}}, then the Hessian matrix is positive-semidefinite at {{mvar|p}}.
The set of positive definite matrices is an [[Open set|open]] [[convex cone]], while the set of positive semi-definition matrices is a [[Closed set|closed]] convex cone.<ref>{{Cite book |last=Boyd |first=Stephen |url=http://dx.doi.org/10.1017/cbo9780511804441 |title=Convex Optimization |last2=Vandenberghe |first2=Lieven |date=2004-03-08 |publisher=Cambridge University Press |isbn=978-0-521-83378-3}}</ref>
Some authors use more general definitions of definiteness, including some non-symmetric real matrices, or non-Hermitian complex ones.
| |||