faster algorithms for the characteristic polynomial
play

Faster algorithms for the characteristic polynomial Clment P ERNET - PowerPoint PPT Presentation

Jun 10, 2023 •308 likes •935 views

State of the art A new algorithm The new algorithm into practice Faster algorithms for the characteristic polynomial Clment P ERNET and Arne S TORJOHANN Symbolic Computation Group University of Waterloo, Canada. ISSAC 2007, Waterloo, July

  1. State of the art A new algorithm The new algorithm into practice Faster algorithms for the characteristic polynomial Clément P ERNET and Arne S TORJOHANN Symbolic Computation Group University of Waterloo, Canada. ISSAC 2007, Waterloo, July 30 C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  2. State of the art A new algorithm The new algorithm into practice Problem Compute the characteristic polynomial of a dense matrix over a field C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  3. State of the art A new algorithm The new algorithm into practice Problem Compute the characteristic polynomial of a dense matrix over a field Result Randomized Las-Vegas algorithm in O ( n ω ) field operations for large fields ( # F > 2 n 2 ). C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  4. State of the art A new algorithm The new algorithm into practice Problem Compute the characteristic polynomial of a dense matrix over a field Result Randomized Las-Vegas algorithm in O ( n ω ) field operations for large fields ( # F > 2 n 2 ). Improves previous complexity by a log n factor, Optimal reduction to Matrix multiplication. C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  5. State of the art A new algorithm The new algorithm into practice Problem Compute the characteristic polynomial of a dense matrix over a field Result Randomized Las-Vegas algorithm in O ( n ω ) field operations for large fields ( # F > 2 n 2 ). Improves previous complexity by a log n factor, Optimal reduction to Matrix multiplication. Practical efficiency. E.g. over Z 547 909 : n 500 5000 15 000 0.91s 4m44s 2h20m LinBox 1.27s 15m32s 7h28m magma-2.13 C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  6. State of the art A new algorithm The new algorithm into practice Outline State of the art 1 A new algorithm 2 Shifted forms Principle of the new algorithm Complexity 3 The new algorithm into practice C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  7. State of the art A new algorithm The new algorithm into practice Outline State of the art 1 A new algorithm 2 Shifted forms Principle of the new algorithm Complexity 3 The new algorithm into practice C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  8. State of the art A new algorithm The new algorithm into practice Pre-Strassen age Leverrier 1840: trace of powers of A , and Newton’s formula improved/rediscovered by Souriau, Faddeev, Frame and Csanky n 4 � � O , based on Matrix multiplication Suited for parallel computation model C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  9. State of the art A new algorithm The new algorithm into practice Pre-Strassen age Leverrier 1840: trace of powers of A , and Newton’s formula improved/rediscovered by Souriau, Faddeev, Frame and Csanky n 4 � � O , based on Matrix multiplication Suited for parallel computation model Danilevskii 1937: elementary row/column operations � n 3 � ⇒O C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  10. State of the art A new algorithm The new algorithm into practice Pre-Strassen age Leverrier 1840: trace of powers of A , and Newton’s formula improved/rediscovered by Souriau, Faddeev, Frame and Csanky n 4 � � O , based on Matrix multiplication Suited for parallel computation model Danilevskii 1937: elementary row/column operations � n 3 � ⇒O Hessenberg 1942: transformation to quasi-upper triangular and determinant expansion formula. n 3 � � ⇒O C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  11. State of the art A new algorithm The new algorithm into practice Post-Strassen age Preparata & Sarwate 1978: Update Csanky with fast matrix multiplication n ω + 1 � � ⇒O C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  12. State of the art A new algorithm The new algorithm into practice Post-Strassen age Preparata & Sarwate 1978: Update Csanky with fast matrix multiplication n ω + 1 � � ⇒O Keller-Gehrig 1985, alg.1: computes ( A 2 i ) i = 1 ... log 2 n to form a Krylov basis. O ( n ω log n ) the best complexity up to now C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  13. State of the art A new algorithm The new algorithm into practice Post-Strassen age Preparata & Sarwate 1978: Update Csanky with fast matrix multiplication n ω + 1 � � ⇒O Keller-Gehrig 1985, alg.1: computes ( A 2 i ) i = 1 ... log 2 n to form a Krylov basis. O ( n ω log n ) the best complexity up to now Keller-Gehrig 1985, alg.2: inspired by Danilevskii, block operations O ( n ω ) but only valid with generic matrices C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  14. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Outline State of the art 1 A new algorithm 2 Shifted forms Principle of the new algorithm Complexity 3 The new algorithm into practice C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  15. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Definition (degree d Krylov matrix of one vector v ) A d − 1 v � � K = v Av . . . Property 2 3 0 ∗ 1 6 ∗ 7 A × K = K × 6 7 ... 6 7 4 5 ∗ 1 ∗ | {z } C PA , v min C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  16. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Definition (degree d Krylov matrix of one vector v ) A d − 1 v � � K = v Av . . . Property 2 3 0 ∗ 1 6 ∗ 7 A × K = K × 6 7 ... 6 7 4 5 ∗ 1 ∗ | {z } C PA , v min ⇒ if d = n , K − 1 AK = C P A car C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  17. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Definition (degree d Krylov matrix of one vector v ) A d − 1 v � � K = v Av . . . Property 2 3 0 ∗ 1 6 ∗ 7 A × K = K × 6 7 ... 6 7 4 5 ∗ 1 ∗ | {z } C PA , v min ⇒ if d = n , K − 1 AK = C P A car [Keller-Gehrig, alg. 2] : K − 1 AK in O ( n ω ) for A generic C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  18. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Definition (degree k Krylov matrix of several vectors v i ) � A k − 1 v 1 A k − 1 v 2 A k − 1 v l � K = v 1 . . . v 2 . . . . . . v l . . . Property k k ≤ k 0 1 1 0 1 1 A × K = K × 0 1 1 C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  19. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Fact (Shift Hessenberg form) If ( d 1 , . . . d l ) is lexicographically maximal such that � A d 1 − 1 v 1 A d l − 1 v l � K = v 1 . . . . . . v l . . . is non-singular, then d 1 d 2 dl 0 1 1 0 1 A × K = K × 1 0 1 1 C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  20. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Principle k -shifted form: k k ≤ k 0 1 1 0 1 1 0 1 1 C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  21. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Principle k + 1-shifted form: k + 1 k + 1 d l 0 1 1 0 1 1 0 1 1 C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  22. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Principle Compute iteratively from 1-shifted form to d 1 -shifted form C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  23. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Principle Compute iteratively from 1-shifted form to d 1 -shifted form each completed block appears in the increasing degree order C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  24. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Principle Compute iteratively from 1-shifted form to d 1 -shifted form each completed block appears in the increasing degree order until the shifted Hessenberg form is obtained: d 1 d 2 d l 0 1 1 0 1 1 0 1 1 C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

  25. State of the art Shifted forms A new algorithm Principle of the new algorithm The new algorithm into practice Complexity Example C. P ERNET and A. S TORJOHANN Faster algorithms for the characteristic polynomial

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Factoring the Characteristic Polynomial Joshua Hallam and Bruce Sagan Department of Mathematics

Factoring the Characteristic Polynomial Joshua Hallam and Bruce Sagan Department of Mathematics

Factoring the Characteristic Polynomial Joshua Hallam and Bruce Sagan Department of Mathematics Michigan State University East Lansing, MI 48824-1027 www.math.msu.edu/ sagan June 21, 2014 Motivating Example Quotient Posets The Standard

1.73k views • 147 slides
Factoring the Characteristic Polynomial Joshua Hallam and Bruce Sagan Department of Mathematics

Factoring the Characteristic Polynomial Joshua Hallam and Bruce Sagan Department of Mathematics

Factoring the Characteristic Polynomial Joshua Hallam and Bruce Sagan Department of Mathematics Michigan State University East Lansing, MI 48824-1027 www.math.msu.edu/ sagan February 20, 2019 Motivating Example Quotient Posets The

673 views • 20 slides
Halt return result; } true 5 6 1 The Halting Problem Undecidability Alan Turing, 1936

Halt return result; } true 5 6 1 The Halting Problem Undecidability Alan Turing, 1936

Polynomial-time Algorithms Most of the algorithms we have seen so far have been polynomial-time algorithms Introduction to NP Completeness Input size n => worst-case running time of n k , Chapter 9 where k is a constant CPTR 318 Can

212 views • 5 slides
A heuristic quasi-polynomial algorithm for discrete logarithm in small characteristic Razvan

A heuristic quasi-polynomial algorithm for discrete logarithm in small characteristic Razvan

ECC, Chennai October 8, 2014 A heuristic quasi-polynomial algorithm for discrete logarithm in small characteristic Razvan Barbulescu 1 Pierrick Gaudry 2 Antoine Joux 3 e 2 Emmanuel Thom IMJ-PRG, Paris Loria, Nancy LIP6, Paris R.

606 views • 41 slides
Advanced Algorithms COMS31900 Approximation algorithms part four Asymptotic Polynomial Time

Advanced Algorithms COMS31900 Approximation algorithms part four Asymptotic Polynomial Time

Advanced Algorithms COMS31900 Approximation algorithms part four Asymptotic Polynomial Time Approximation Schemes Rapha el Clifford Slides by Benjamin Sach Approximation Algorithms Recap A polynomial time algorithm A is an

1.52k views • 134 slides
NP-Completeness 1 Almost all the algorithms we have studies so far are polynomial-time

NP-Completeness 1 Almost all the algorithms we have studies so far are polynomial-time

Chapter 5 NP-Completeness 1 Almost all the algorithms we have studies so far are polynomial-time algorithms, i.e., the running time is O ( n k ) for some constant k . Generally, we think of problems that are solvable by polynomial-time

532 views • 42 slides
Rounding and Chaining LLL: Finding Faster Small Roots of Univariate Polynomial Congruences J.

Rounding and Chaining LLL: Finding Faster Small Roots of Univariate Polynomial Congruences J.

Rounding and Chaining LLL: Finding Faster Small Roots of Univariate Polynomial Congruences J. Bi, J-S. Coron, J-C. Faug` ere, P . Nguyen, G. Renault, R. Zeitoun Public Key Cryptography 2014 26-28 March, 2014 - Buenos Aires, Argentina

613 views • 36 slides
Faster Bootstrapping with Polynomial Error Jacob Alperin-Sheriff Chris Peikert School of

Faster Bootstrapping with Polynomial Error Jacob Alperin-Sheriff Chris Peikert School of

Faster Bootstrapping with Polynomial Error Jacob Alperin-Sheriff Chris Peikert School of Computer Science Georgia Tech CRYPTO 2014 19 August 2014 1 / 10 Fully Homomorphic Encryption [RAD78,Gentry09] FHE lets you do this: Eval (

1.03k views • 50 slides
MA/CSSE 473 Day 38 Problems Decision Problems P and NP Polynomial time algorithms INTRO TO

MA/CSSE 473 Day 38 Problems Decision Problems P and NP Polynomial time algorithms INTRO TO

MA/CSSE 473 Day 38 Problems Decision Problems P and NP Polynomial time algorithms INTRO TO COMPUTATIONAL COMPLEXITY 1 The Law of the Algorithm Jungle Polynomial good, exponential bad! The latter is obvious, the former may need some

510 views • 16 slides
Advanced Algorithms COMS31900 Approximation algorithms part three (Fully) Polynomial Time

Advanced Algorithms COMS31900 Approximation algorithms part three (Fully) Polynomial Time

Advanced Algorithms COMS31900 Approximation algorithms part three (Fully) Polynomial Time Approximation Schemes Rapha el Clifford Slides by Benjamin Sach Approximation Algorithms Recap An algorithm A is an -approximation algorithm for

1.72k views • 161 slides
Randomized Algorithms Polynomial Identity Testing (PIT) Input : of

Randomized Algorithms Polynomial Identity Testing (PIT) Input : of

Randomized Algorithms Polynomial Identity Testing (PIT) Input : of degree d f, g F [ x 1 , x 2 , . . . , x n ] Output : f g ? F [ x 1 , x 2 , . . . , x n ] : ring of n -variate polynomials over field F f F [ x

923 views • 32 slides
On the maximum of the characteristic polynomial of the Circular Beta Ensemble Joseph Najnudel

On the maximum of the characteristic polynomial of the Circular Beta Ensemble Joseph Najnudel

Presentation of the setting Statement of the main result Orthogonal polynomial on the unit circle Sketch of proof of a non-sharp upper bound Sketch of proof of a sharper upper bound Strategy for a lower bound On the maximum of the

930 views • 63 slides
Singularities in characteristic zero and singularities in characteristic p Karl Schwede 1 1

Singularities in characteristic zero and singularities in characteristic p Karl Schwede 1 1

Singularities on algebraic varieties Types of singularities in characteristic zero Singularities in characteristic p > 0 Singularities in characteristic zero and singularities in characteristic p Karl Schwede 1 1 Department of Mathematics

1.76k views • 146 slides
Faster Algorithms for the Constrained k -means Problem Ragesh Jaiswal CSE, IIT Delhi June 16,

Faster Algorithms for the Constrained k -means Problem Ragesh Jaiswal CSE, IIT Delhi June 16,

Faster Algorithms for the Constrained k -means Problem Ragesh Jaiswal CSE, IIT Delhi June 16, 2015 [Joint work with Anup Bhattacharya (IITD) and Amit Kumar (IITD)] Ragesh Jaiswal Faster Algorithms for the Constrained k -means Problem k -means

664 views • 52 slides
Polynomial-Time Problems Lecture 4: The polynomial method Part II: All-pairs shortest paths

Polynomial-Time Problems Lecture 4: The polynomial method Part II: All-pairs shortest paths

Complexity Theory of Polynomial-Time Problems Lecture 4: The polynomial method Part II: All-pairs shortest paths Sebastian Krinninger Overview on APSP Algorithms Floyd-Warshall algorithm: ( 3 ) Inserts one node at a time

687 views • 42 slides
GUI Testing Chapter 19 GUI characteristic Figure 19.1 What is the main characteristic of

GUI Testing Chapter 19 GUI characteristic Figure 19.1 What is the main characteristic of

GUI Testing Chapter 19 GUI characteristic Figure 19.1 What is the main characteristic of GUIs? GUI2 GUI characteristic 2 What is the main characteristic of GUIs? Event driven GUI3 Levels of testing What are the

718 views • 43 slides
Eigenvalues, Eigenvectors, and Diagonalization Diagonalization Math 240 Calculus III Summer

Eigenvalues, Eigenvectors, and Diagonalization Diagonalization Math 240 Calculus III Summer

Eigenvalues, Eigenvectors, and Diagonal- ization Math 240 Eigenvalues and Eigenvectors Eigenvalues, Eigenvectors, and Diagonalization Diagonalization Math 240 Calculus III Summer 2015, Session II Thursday, July 16, 2015 Eigenvalues,

171 views • 15 slides
Eigenvalues and Eigenvectors Raibatak Sen Gupta 2019 Eigenvalues Characteristic Equation and

Eigenvalues and Eigenvectors Raibatak Sen Gupta 2019 Eigenvalues Characteristic Equation and

Eigenvalues and Eigenvectors Raibatak Sen Gupta Eigenvalues and Eigenvectors Raibatak Sen Gupta 2019 Eigenvalues Characteristic Equation and Eigenvectors Raibatak Sen Gupta Let A be an n n matrix. Then det ( A xI n ) gives a

355 views • 24 slides
Magnetohydrodynamic Turbulence Wolf-Christian Mller Max-Planck-Institut fr Plasmaphysik,

Magnetohydrodynamic Turbulence Wolf-Christian Mller Max-Planck-Institut fr Plasmaphysik,

Magnetohydrodynamic Turbulence Wolf-Christian Mller Max-Planck-Institut fr Plasmaphysik, Garching, Germany Dieter Biskamp, Max-Planck-Institut fr Plasmaphysik, Garching, Germany Roland Grappin, Observatoire de Paris-Meudon, Meudon, France

607 views • 36 slides
Bond Accountability Committee Meeting July 18, 2018 July 2018 July 2018 1 Agenda Welcome

Bond Accountability Committee Meeting July 18, 2018 July 2018 July 2018 1 Agenda Welcome

School Improvement Bond Update Bond Accountability Committee Meeting July 18, 2018 July 2018 July 2018 1 Agenda Welcome & Introductions Public Comments Program Update Project Reports BAC Charter and Responsibility

830 views • 42 slides
Circulant Matrices and Polynomials Dave Frank What is a

Circulant Matrices and Polynomials Dave Frank What is a

1/25 Circulant Matrices and Polynomials Dave Frank What is a Circulant Matrix? An n n circulant matrix is formed by starting with a vector with n 2/25 components. This vector becomes the first row of the

516 views • 25 slides
Numerical Rootfinding in a Compact Region Suzanna Stephenson Brigham Young University January

Numerical Rootfinding in a Compact Region Suzanna Stephenson Brigham Young University January

Numerical Rootfinding in a Compact Region Suzanna Stephenson Brigham Young University January 26, 2019 Introduction Optimization problems are extremely important Many problems have lots of variables, especially in machine learning Most

307 views • 30 slides
ode ode Basic Concepts and Theorems The n th order linear ODE takes the form: n n 1 d y

ode ode Basic Concepts and Theorems The n th order linear ODE takes the form: n n 1 d y

So Solution tion of of hi high gher er or orde der r ode ode Basic Concepts and Theorems The n th order linear ODE takes the form: n n 1 d y d y dy a a ... a a y f x ( ) n n 1 1

440 views • 18 slides
Characteristic Modes Part I: Introduction Miloslav Capek Department of Electromagnetic Field

Characteristic Modes Part I: Introduction Miloslav Capek Department of Electromagnetic Field

Characteristic Modes Part I: Introduction Miloslav Capek Department of Electromagnetic Field Czech Technical University in Prague, Czech Republic miloslav.capek@fel.cvut.cz Seminar May 2, 2018 Capek, M. Characteristic Modes Part

832 views • 82 slides

More recommend