12.07.2015 Views

v2007.09.17 - Convex Optimization

v2007.09.17 - Convex Optimization

v2007.09.17 - Convex Optimization

SHOW MORE
SHOW LESS
  • No tags were found...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

542 APPENDIX C. SOME ANALYTICAL OPTIMAL RESULTSFor B ∈ S N whose eigenvalues λ(B)∈ R N are arranged in nonincreasingorder, and for diagonal matrix Υ∈ S k whose diagonal entries arearranged in nonincreasing order where 1≤k ≤N , we utilize themain-diagonal δ operator’s self-adjointness property (1223): [10,4.2]k∑i=1k∑i=1Υ ii λ(B) N−i+1 = inf tr(ΥU T BU) = infU∈ R N×kU T U=I= minimizeV i ∈S Ni=1U∈ R N×kU T U=Iδ(Υ) T δ(U T BU)()∑tr B k (Υ ii −Υ i+1,i+1 )V isubject to trV i = i ,I ≽ V i ≽ 0,where Υ k+1,k+1 ∆ = 0. We speculate,k∑Υ ii λ(B) i = supi=1U∈ R N×kU T U=IAlizadeh shows: [9,4.2]tr(ΥU T BU) = supU∈ R N×kU T U=Ii=1... ki=1... k(1481)δ(Υ) T δ(U T BU) (1482)Υ ii λ(B) i = minimizek∑iµ i + trZ iµ∈R k , Z i ∈S N i=1subject to µ i I + Z i − (Υ ii −Υ i+1,i+1 )B ≽ 0, i=1... k= maximizeV i ∈S Nwhere Υ k+1,k+1 ∆ = 0.Z i ≽ 0,()∑tr B k (Υ ii −Υ i+1,i+1 )V ii=1i=1... ksubject to trV i = i , i=1... kI ≽ V i ≽ 0, i=1... k (1483)The largest eigenvalue magnitude µ of A∈ S Nmax { |λ(A) i | } = minimize µi µ∈Rsubject to −µI ≼ A ≼ µI(1484)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!