solvability complexity index sci and towers of algorithms
play

Solvability Complexity Index (=SCI) and Towers of Algorithms Olavi - PowerPoint PPT Presentation

Solvability Complexity Index (=SCI) and Towers of Algorithms Olavi Nevanlinna Aalto SCI February 20, 2015 Goal of the talk Is ( A ) computable for A B ( 2 ( N )) Goal of the talk Is ( A ) computable for A B ( 2 (


  1. Solvability Complexity Index (=SCI) and Towers of Algorithms Olavi Nevanlinna Aalto SCI February 20, 2015

  2. Goal of the talk ◮ Is σ ( A ) computable for A ∈ B ( ℓ 2 ( N ))

  3. Goal of the talk ◮ Is σ ( A ) computable for A ∈ B ( ℓ 2 ( N )) ◮ To explain what different theories say about it

  4. Goal of the talk ◮ Is σ ( A ) computable for A ∈ B ( ℓ 2 ( N )) ◮ To explain what different theories say about it ◮ This is a simplified layman overview

  5. Goal of the talk ◮ Is σ ( A ) computable for A ∈ B ( ℓ 2 ( N )) ◮ To explain what different theories say about it ◮ This is a simplified layman overview ◮ Then I focus on Towers of Algorithms and on the Solvability Complexity Index,

  6. Goal of the talk ◮ Is σ ( A ) computable for A ∈ B ( ℓ 2 ( N )) ◮ To explain what different theories say about it ◮ This is a simplified layman overview ◮ Then I focus on Towers of Algorithms and on the Solvability Complexity Index, ◮ J. Ben-Artzi, A. Hansen, O. Nevanlinna , M. Seidel

  7. Definition of a Tower PROBLEM B ( ℓ 2 ( N )) Ω: primary set, e.g Λ: evaluation set, e.g. f ij : A �→ < Ae i , e j > for A ∈ B ( ℓ 2 ( N )) M : metric space Ξ: problem function Ω → M , such as σ ( A ) for A ∈ B ( ℓ 2 ( N )) TOWER Ξ( A ) = lim n k →∞ Γ n k ( A ) Γ n k ( A ) := lim n k − 1 →∞ Γ n k , n k − 1 ( A ) ..... ..... ..... Γ n k ,., n 2 ( A ) := lim n 1 →∞ Γ n k ,., n 2 , n 1 ( A )

  8. Matrices first A ∈ B ( C n ) solve for π A ( z ) = 0 ◮ n ≤ 3 : generally convergent rational iteration exists (McMullen 1987)

  9. Matrices first A ∈ B ( C n ) solve for π A ( z ) = 0 ◮ n ≤ 3 : generally convergent rational iteration exists (McMullen 1987) ◮ n ≤ 5 : a tower of generally convergent rational iterations (Doyle, McMullen 1989)

  10. Matrices first A ∈ B ( C n ) solve for π A ( z ) = 0 ◮ n ≤ 3 : generally convergent rational iteration exists (McMullen 1987) ◮ n ≤ 5 : a tower of generally convergent rational iterations (Doyle, McMullen 1989) ◮ n > 5 : no such towers (Doyle, McMullen 1989)

  11. Matrices continues radicals, z �→ | z | available, then convergent iterations exist for solving roots of polynomials input finite: the complex coefficients of the polynomial

  12. Computabilities... ”Turing view”: problem computable if a computing device exists which solves the problem Computation in the limit and higher hierarchies BSS (Blum, Shub, Smale) R -machine model IBC (infromation based complexity) uses BSS, ”tractability” constructivism, computability on Z and within computable numbers

  13. Any compact can be spectrum Represent compact K ⊂ C from outside: � K = K n where · · · ⊂ K n +1 ⊂ K n ⊂ · · · and testing z / ∈ K n ”easy”

  14. Any compact can be spectrum, so look at Julia sets We first look at the Julia set J for the quadratic polynomial z 2 + 4. Consider the question z ∈ J ? Then the corresponding question for the spectrum σ ( A ) is λ ∈ σ ( A ) ? The natural formulation of these questions is, can you decide whether the answer is yes or no?

  15. 2.1 Julia set J for z 2 + 4 Let p ( z ) = z 2 + 4 Iterate z n +1 = p ( z n ) If z n → ∞ then z 0 / ∈ J . √ Note that if | z k | > 1 + 5 for some k , then | z k +1 | > 2 | z k | and then z n → ∞ . For this p ( z ) the Julia set is homeomorphic to a Cantor set. Observe that C \ J is open. S. Smale and coworkers: J is not decidable (”semidecidable”)

  16. Computation in the limit... Output as follows: √ if | z k | ≤ 1 + 5 , then Out ( k ) = 1 √ if | z k | > 1 + 5, then Out ( k ) = 0. So depending on the initial value we obtain sequences of the form 1 , 1 , . . . , 1 , 0 , 0 , 0 . . . and 1 , 1 , 1 , . . . In either case the limit exists; and then you (would) know

  17. Similar question for the spectrum in abstract Banach algebra Consider the subalgebra generated by just one element a (say, in Banach algebra A ). Then the spectrum within the subalgebra is fill ( σ ( a )). If we are allowed to produce polynomials of a and compute their norms but inverting is not allowed, then: The question ∈ fill ( σ ( a )) λ / is semidecidable as follows: If answer positive: finite termination with sure answer, while if negative, you will never know (the one you look after does not exist)

  18. What exists is easier to find! Conclude: Think positive, construct the resolvent C \ fill ( σ ( A )) → B ( X ) λ �→ ( λ − A ) − 1 instead! Get a multicentric holomorphic calculus - but not during this talk...

  19. Computation in the limit Example Let A be diagonal operator in ℓ 2 ( N ) such that a ii ∈ { 0 , 1 } . Input information: read one diagonal element in time, in a fixed enumeration. Then ◮ σ ( A ) ∈ { 0 , 1 } : this we can build in the ”machine” based on the problem description

  20. Computation in the limit Example Let A be diagonal operator in ℓ 2 ( N ) such that a ii ∈ { 0 , 1 } . Input information: read one diagonal element in time, in a fixed enumeration. Then ◮ σ ( A ) ∈ { 0 , 1 } : this we can build in the ”machine” based on the problem description ◮ σ ess ( A ) � = ∅ : this can also be build in

  21. Computation in the limit Example Let A be diagonal operator in ℓ 2 ( N ) such that a ii ∈ { 0 , 1 } . Input information: read one diagonal element in time, in a fixed enumeration. Then ◮ σ ( A ) ∈ { 0 , 1 } : this we can build in the ”machine” based on the problem description ◮ σ ess ( A ) � = ∅ : this can also be build in ◮ 1 ∈ σ ( A ): this cannot be be computed except at the limit

  22. Computation in the limit Example Let A be diagonal operator in ℓ 2 ( N ) such that a ii ∈ { 0 , 1 } . Input information: read one diagonal element in time, in a fixed enumeration. Then ◮ σ ( A ) ∈ { 0 , 1 } : this we can build in the ”machine” based on the problem description ◮ σ ess ( A ) � = ∅ : this can also be build in ◮ 1 ∈ σ ( A ): this cannot be be computed except at the limit ◮ 1 ∈ σ ess ( A ) this needs ”two limits”, i.e. a ”tower”

  23. How to get the answers 1 ∈ σ ( A ) ◮ define function for each n n � Γ n ( A ) = 1 , if a ii > 0 , i =1 0 , otherwise and set Γ( A ) = lim n →∞ Γ n ( A ) . Then, answer is ”yes”, when Γ( A ) = 1

  24. How to get the answers 1 ∈ σ ( A ) ◮ define function for each n n � Γ n ( A ) = 1 , if a ii > 0 , i =1 0 , otherwise and set Γ( A ) = lim n →∞ Γ n ( A ) . Then, answer is ”yes”, when Γ( A ) = 1 ◮ Using quantifiers: ∃ n ( � n i =1 a ii > 0)

  25. How to get the answers 1 ∈ σ ess ( A ) ◮ this needs ”two limits”, i.e. a ”tower” of height 2 n � Γ m , n ( A ) = 1 , if a ii > m , i =1 0 , otherwise Γ m ( A ) = lim n →∞ Γ m , n ( A ) Γ( A ) = lim m →∞ Γ m ( A ) Again, answer is ”yes”, when Γ( A ) = 1

  26. How to get the answers 1 ∈ σ ess ( A ) ◮ this needs ”two limits”, i.e. a ”tower” of height 2 n � Γ m , n ( A ) = 1 , if a ii > m , i =1 0 , otherwise Γ m ( A ) = lim n →∞ Γ m , n ( A ) Γ( A ) = lim m →∞ Γ m ( A ) Again, answer is ”yes”, when Γ( A ) = 1 ◮ With two quantifiers: ∀ m ∃ n ( � n i =1 a ii > m )

  27. Another example We define A ∈ B ( ℓ 2 ( N )) using diagonal blocks: ∞ � A = A k ( j ) j =1 where A k are k × k -matrices with number 1’s in the corners, like   1 0 1 A 3 = 0 0 0   1 0 1 and k ( j ) ≥ 2 is some sequence. Thus, A is algebraic, σ ( A ) = σ ess ( A ) = { 0 , 2 } .

  28. Constructivism, computability ◮ The operator ∞ � A = A k ( j ) j =1 is effectively determined if one can determine the sequence { k ( j ) } recursively.

  29. Constructivism, computability ◮ The operator ∞ � A = A k ( j ) j =1 is effectively determined if one can determine the sequence { k ( j ) } recursively. ◮ But,

  30. Constructivism, computability ◮ The operator ∞ � A = A k ( j ) j =1 is effectively determined if one can determine the sequence { k ( j ) } recursively. ◮ But, ◮ then one can ”tailor” a computing machine which computes the spectrum in a finite number of operations

  31. Constructivism, computability 2 ◮ The operator ∞ � B = β j A k ( j ) j =1 is effectively determined if one can determine the sequence { k ( j ) } recursively and the coefficient sequence { β j } is a computable sequence of reals.

  32. Constructivism, computability 2 ◮ The operator ∞ � B = β j A k ( j ) j =1 is effectively determined if one can determine the sequence { k ( j ) } recursively and the coefficient sequence { β j } is a computable sequence of reals. ◮ Then,

  33. Constructivism, computability 2 ◮ The operator ∞ � B = β j A k ( j ) j =1 is effectively determined if one can determine the sequence { k ( j ) } recursively and the coefficient sequence { β j } is a computable sequence of reals. ◮ Then, ◮ the spectrum is computable.

  34. Constructivism, computability 3 ◮ In this theory effectively described bounded self-adjoint operators have computable spectra

  35. Constructivism, computability 3 ◮ In this theory effectively described bounded self-adjoint operators have computable spectra ◮ but

Recommend


More recommend