v2009.01.01 - Convex Optimization
v2009.01.01 - Convex Optimization v2009.01.01 - Convex Optimization
532 CHAPTER 7. PROXIMITY PROBLEMS dimension not in excess of that ρ desired; id est, spectral projection on ⎡ ⎤ ⎣ Rρ+1 + 0 ⎦ ∩ ∂H ⊂ R N+1 (1306) R − where ∂H = {λ ∈ R N+1 | 1 T λ = 0} (1017) is a hyperplane through the origin. This pointed polyhedral cone (1306), to which membership subsumes the rank constraint, has empty interior. Given desired affine dimension 0 ≤ ρ ≤ N −1 and diagonalization (A.5) of unknown EDM D [ 0 1 T 1 −D ] ∆ = UΥU T ∈ S N+1 h (1307) and given symmetric H in diagonalization [ ] 0 1 T ∆ = QΛQ T ∈ S N+1 (1308) 1 −H having eigenvalues arranged in nonincreasing order, then by (1030) problem (1304) is equivalent to ∥ minimize ∥δ(Υ) − π ( δ(R T ΛR) )∥ ∥ 2 Υ , R ⎡ ⎤ subject to δ(Υ) ∈ ⎣ Rρ+1 + 0 ⎦ ∩ ∂H (1309) R − δ(QRΥR T Q T ) = 0 R −1 = R T where π is the permutation operator from7.1.3 arranging its vector argument in nonincreasing order, 7.18 where R ∆ = Q T U ∈ R N+1×N+1 (1310) in U on the set of orthogonal matrices is a bijection, and where ∂H insures one negative eigenvalue. Hollowness constraint δ(QRΥR T Q T ) = 0 makes problem (1309) difficult by making the two variables dependent. 7.18 Recall, any permutation matrix is an orthogonal matrix.
7.3. THIRD PREVALENT PROBLEM: 533 Our plan is to instead divide problem (1309) into two and then iterate their solution: ∥ minimize ∥δ(Υ) − π ( δ(R T ΛR) )∥ ∥ 2 Υ ⎡ ⎤ subject to δ(Υ) ∈ ⎣ Rρ+1 + (a) 0 ⎦ ∩ ∂H R − (1311) minimize ‖R Υ ⋆ R T − Λ‖ 2 F R subject to δ(QR Υ ⋆ R T Q T ) = 0 R −1 = R T (b) Proof. We justify disappearance of the hollowness constraint in convex optimization problem (1311a): From the arguments in7.1.3 with regard ⎡ to⎤π the permutation operator, cone membership constraint δ(Υ) ∈⎣ Rρ+1 + 0 ⎦∩ ∂H from (1311a) is equivalent to R − ⎡ ⎤ δ(Υ) ∈ ⎣ Rρ+1 + 0 ⎦ ∩ ∂H ∩ K M (1312) R − where K M is the monotone cone (2.13.9.4.3). Membership of δ(Υ) to the polyhedral cone of majorization (Theorem A.1.2.0.1) K ∗ λδ = ∂H ∩ K ∗ M+ (1329) where K ∗ M+ is the dual monotone nonnegative cone (2.13.9.4.2), is a condition (in absence of a hollowness [ constraint) ] that would insure existence 0 1 T of a symmetric hollow matrix . Curiously, intersection of 1 −D ⎡ ⎤ this feasible superset ⎣ Rρ+1 + 0 ⎦∩ ∂H ∩ K M from (1312) with the cone of majorization K ∗ λδ R − is a benign operation; id est, ∂H ∩ K ∗ M+ ∩ K M = ∂H ∩ K M (1313)
- Page 481 and 482: 6.8. DUAL EDM CONE 481 EDM 2 = S 2
- Page 483 and 484: 6.8. DUAL EDM CONE 483 whose veraci
- Page 485 and 486: 6.8. DUAL EDM CONE 485 6.8.1.3.1 Ex
- Page 487 and 488: 6.8. DUAL EDM CONE 487 has dual aff
- Page 489 and 490: 6.8. DUAL EDM CONE 489 6.8.1.7 Scho
- Page 491 and 492: 6.9. THEOREM OF THE ALTERNATIVE 491
- Page 493 and 494: 6.10. POSTSCRIPT 493 When D is an E
- Page 495 and 496: Chapter 7 Proximity problems In sum
- Page 497 and 498: 497 project on the subspace, then p
- Page 499 and 500: 499 H S N h 0 EDM N K = S N h ∩ R
- Page 501 and 502: 501 7.0.3 Problem approach Problems
- Page 503 and 504: 7.1. FIRST PREVALENT PROBLEM: 503 f
- Page 505 and 506: 7.1. FIRST PREVALENT PROBLEM: 505 7
- Page 507 and 508: 7.1. FIRST PREVALENT PROBLEM: 507 d
- Page 509 and 510: 7.1. FIRST PREVALENT PROBLEM: 509 7
- Page 511 and 512: 7.1. FIRST PREVALENT PROBLEM: 511 w
- Page 513 and 514: 7.1. FIRST PREVALENT PROBLEM: 513 T
- Page 515 and 516: 7.2. SECOND PREVALENT PROBLEM: 515
- Page 517 and 518: 7.2. SECOND PREVALENT PROBLEM: 517
- Page 519 and 520: 7.2. SECOND PREVALENT PROBLEM: 519
- Page 521 and 522: 7.2. SECOND PREVALENT PROBLEM: 521
- Page 523 and 524: 7.2. SECOND PREVALENT PROBLEM: 523
- Page 525 and 526: 7.3. THIRD PREVALENT PROBLEM: 525 g
- Page 527 and 528: 7.3. THIRD PREVALENT PROBLEM: 527 w
- Page 529 and 530: 7.3. THIRD PREVALENT PROBLEM: 529 7
- Page 531: 7.3. THIRD PREVALENT PROBLEM: 531 7
- Page 535 and 536: 7.4. CONCLUSION 535 The rank constr
- Page 537 and 538: Appendix A Linear algebra A.1 Main-
- Page 539 and 540: A.1. MAIN-DIAGONAL δ OPERATOR, λ
- Page 541 and 542: A.1. MAIN-DIAGONAL δ OPERATOR, λ
- Page 543 and 544: A.2. SEMIDEFINITENESS: DOMAIN OF TE
- Page 545 and 546: A.3. PROPER STATEMENTS 545 (AB) T
- Page 547 and 548: A.3. PROPER STATEMENTS 547 A.3.1 Se
- Page 549 and 550: A.3. PROPER STATEMENTS 549 For A di
- Page 551 and 552: A.3. PROPER STATEMENTS 551 Diagonal
- Page 553 and 554: A.3. PROPER STATEMENTS 553 For A,B
- Page 555 and 556: A.3. PROPER STATEMENTS 555 A.3.1.0.
- Page 557 and 558: A.4. SCHUR COMPLEMENT 557 A.4 Schur
- Page 559 and 560: A.4. SCHUR COMPLEMENT 559 A.4.0.0.2
- Page 561 and 562: A.5. EIGEN DECOMPOSITION 561 When B
- Page 563 and 564: A.5. EIGEN DECOMPOSITION 563 dim N(
- Page 565 and 566: A.6. SINGULAR VALUE DECOMPOSITION,
- Page 567 and 568: A.6. SINGULAR VALUE DECOMPOSITION,
- Page 569 and 570: A.7. ZEROS 569 Given symmetric matr
- Page 571 and 572: A.7. ZEROS 571 (TRANSPOSE.) Likewis
- Page 573 and 574: A.7. ZEROS 573 For X,A∈ S M + [31
- Page 575 and 576: A.7. ZEROS 575 A.7.5.0.1 Propositio
- Page 577 and 578: Appendix B Simple matrices Mathemat
- Page 579 and 580: B.1. RANK-ONE MATRIX (DYAD) 579 R(v
- Page 581 and 582: B.1. RANK-ONE MATRIX (DYAD) 581 ran
7.3. THIRD PREVALENT PROBLEM: 533<br />
Our plan is to instead divide problem (1309) into two and then iterate<br />
their solution:<br />
∥<br />
minimize ∥δ(Υ) − π ( δ(R T ΛR) )∥ ∥ 2<br />
Υ ⎡ ⎤<br />
subject to δ(Υ) ∈ ⎣ Rρ+1 +<br />
(a)<br />
0 ⎦ ∩ ∂H<br />
R −<br />
(1311)<br />
minimize ‖R Υ ⋆ R T − Λ‖ 2 F<br />
R<br />
subject to δ(QR Υ ⋆ R T Q T ) = 0<br />
R −1 = R T<br />
(b)<br />
Proof. We justify disappearance of the hollowness constraint in<br />
convex optimization problem (1311a): From the arguments in7.1.3<br />
with regard ⎡ to⎤π the permutation operator, cone membership constraint<br />
δ(Υ) ∈⎣ Rρ+1 +<br />
0 ⎦∩ ∂H from (1311a) is equivalent to<br />
R −<br />
⎡ ⎤<br />
δ(Υ) ∈ ⎣ Rρ+1 +<br />
0 ⎦ ∩ ∂H ∩ K M (1312)<br />
R −<br />
where K M is the monotone cone (2.13.9.4.3). Membership of δ(Υ) to the<br />
polyhedral cone of majorization (Theorem A.1.2.0.1)<br />
K ∗ λδ = ∂H ∩ K ∗ M+ (1329)<br />
where K ∗ M+ is the dual monotone nonnegative cone (2.13.9.4.2), is a<br />
condition (in absence of a hollowness [ constraint) ] that would insure existence<br />
0 1<br />
T<br />
of a symmetric hollow matrix . Curiously, intersection of<br />
1 −D<br />
⎡ ⎤<br />
this feasible superset ⎣ Rρ+1 +<br />
0 ⎦∩ ∂H ∩ K M from (1312) with the cone of<br />
majorization K ∗ λδ<br />
R −<br />
is a benign operation; id est,<br />
∂H ∩ K ∗ M+ ∩ K M = ∂H ∩ K M (1313)