Home
Wikimization
Contact Us
Accumulator error feedback
CVX Download
Calculus of Inequalities
Rick Chartrand
Chromosome Structure EDM
Complementarity problem
Compressive Sampling
Compressed Sensing
Conic Independence
Convex Cones
Convex Functions
Convex Geometry
Convex, Affine, Conic: Hulls
Convex Iteration
Convex Optimization
Convex Optimization Group
Dattorro PC Optimization
Dattorro Supercomputer
Distance Geometry
Distance Matrix Cone
Dual Cones
Duality Gap
Eigenvalues/Eigenvectors
Elliptope and Fantope
Euclidean Distance Matrices
EDM cone faces
Extreme Directions
Face Recognition
Farkas Lemma
Fermat point
Fifth Metric Property
Jensen's Inequality
Jobs in Optimization
Kissing Number
Harold W. Kuhn
Linear Algebra
Linear Matrix Inequality
Manifold Learning
MATLAB for Optimization
Matrix Calculus
Molecular Conformation
Moreau's theorem
Isaac Newton
Angelia Nedic
Open Problems
Positive Matrix Factorization
Positive Semidefinite Cone
Projection
Projection on Cone
Proximity Problems
PY4SCIENCE
Quasiconvex Functions
Rank Constraint
Rockafellar
Justin Romberg
Michael Saunders
Schoenberg Criterion
Semidefinite Programming
Sensor Network Localization
Smallest Simplex
Systems Optimization Lab
Stanford SOL
Talks on Optimization
Joshua Trzasko
Video
Wikimization     Meboo     SOL      Video     CVX     Contact     
Felice crystal
Home arrow Linear Algebra
Linear Algebra

We call  x^T A x >= 0  for all x the most fundamental test of positive semidefiniteness.  Yet some authors instead say, for real A and complex domain, the complex test   x^H A x >= 0   is most fundamental.  That complex broadening of the domain of test causes nonsymmetric real matrices to be excluded from the set of positive semidefinite matrices.  Yet admitting nonsymmetric real matrices or not is a matter of preference unless that complex test is adopted, as we shall now explain.

        sensor network localization

Any real square matrix A has a representation in terms of its symmetric and antisymmetric parts.  Because, for all real A the antisymmetric part vanishes under real test x^T(A-A^T)x=0, only the symmetric part (A+A^T)/2  has a role determining positive semidefiniteness.  Hence the oft-made presumption that only symmetric matrices may be positive semidefinite is, of course, erroneous under the most fundamental test.  Because eigenvalue-signs of a symmetric matrix translate unequivocally to its semidefiniteness, the eigenvalues that determine semidefiniteness are always those of the symmetrized matrix.  For that reason, and because symmetric (or Hermitian) matrices must have real eigenvalues, the convention adopted in the literature is that semidefinite matrices are synonymous with symmetric semidefinite matrices.  Certainly misleading under the most fundamental test, that presumption is typically bolstered with compelling examples from the physical sciences where symmetric matrices occur within the mathematical exposition of natural phenomena.

Read more...

 
Course,   Video
Convex Optimization
     convex optimization
Stephen Boyd 
L. Vandenberghe 


Dattorro      convex optimization Euclidean distance geometry 2ε
Dattorro


Course
Bertsekas
     books by Bertsekas
Dimitri Bertsekas 


See Inside Hiriart-Urruty & Lemaréchal
Hiriart-Urruty
& Lemaréchal


See Inside
Rockafellar Rockafellar