Practical shift choice in the shift-and-invert Krylov subspace evaluations of the matrix exponential

We propose two methods to find a proper shift parameter in the shift-and-invert method for computing matrix exponential matrix-vector products. These methods are useful in the case of matrix exponential action has to be computed for a number of vectors. The first method is based on the zero-order optimization of the mean residual norm for a given number of initial vectors. The second method processes the vectors one-by-one and estimates, for each vector, the derivative of the residual norm as a function of the shift parameter. The estimated derivative value is then used to update the shift value for the next vector. To demonstrate the performance of the proposed methods we perform numerical experiments for two-dimensional non-stationary convection-diffusion equation with discontinuous coefficients and two-dimensional anisotropic diffusion equation.

[1]  Valeria Simoncini,et al.  A new investigation of the extended Krylov subspace method for matrix function evaluations , 2009, Numer. Linear Algebra Appl..

[2]  Awad H. Al-Mohy,et al.  Computing the Action of the Matrix Exponential, with an Application to Exponential Integrators , 2011, SIAM J. Sci. Comput..

[3]  Valeria Simoncini,et al.  Analysis of Projection Methods for Rational Function Approximation to the Matrix Exponential , 2006, SIAM J. Numer. Anal..

[4]  E. Grimme,et al.  Pade approximation of large-scale dynamic systems with Lanczos methods , 1994, Proceedings of 1994 33rd IEEE Conference on Decision and Control.

[5]  Igor Moret,et al.  The restarted shift‐and‐invert Krylov method for matrix functions , 2014, Numer. Linear Algebra Appl..

[6]  Eric Jones,et al.  SciPy: Open Source Scientific Tools for Python , 2001 .

[7]  I. Moret,et al.  RD-Rational Approximations of the Matrix Exponential , 2004 .

[8]  Valeria Simoncini,et al.  Adaptive rational Krylov subspaces for large-scale dynamical systems , 2011, Syst. Control. Lett..

[9]  Marlis Hochbruck,et al.  Preconditioning Lanczos Approximations to the Matrix Exponential , 2005, SIAM J. Sci. Comput..

[10]  Michele Benzi,et al.  On the Limiting Behavior of Parameter-Dependent Network Centrality Measures , 2013, SIAM J. Matrix Anal. Appl..

[11]  W. Stewart,et al.  A numerical study of large sparse matrix exponentials arising in Markov chains 1 1 This work has ben , 1999 .

[12]  Stefan Güttel,et al.  Generalized Rational Krylov Decompositions with an Application to Rational Approximation , 2015, SIAM J. Matrix Anal. Appl..

[13]  C. Kelley.  Iterative Methods for Linear and Nonlinear Equations , 1987 .

[14]  G Deco,et al.  Framework based on communicability and flow to analyze complex network dynamics. , 2018, Physical review. E.

[15]  Lothar Reichel,et al.  Analysis of directed networks via the matrix exponential , 2019, J. Comput. Appl. Math..

[16]  Awad H. Al-Mohy,et al.  A New Scaling and Squaring Algorithm for the Matrix Exponential , 2009, SIAM J. Matrix Anal. Appl..

[17]  A. Katrutsa,et al.  How to optimize preconditioners for the conjugate gradient method: a stochastic approach , 2018, 1806.06045.

[18]  D. Anderson,et al.  Algorithms for minimization without derivatives , 1974 .

[19]  Marlis Hochbruck,et al.  Residual, Restarting, and Richardson Iteration for the Matrix Exponential , 2010, SIAM J. Sci. Comput..

[20]  Solution of the principal spectral problem and mathematical modeling of nuclear reactors , 2000 .

[21]  S. Güttel.  Rational Krylov approximation of matrix functions: Numerical methods and optimal pole selection , 2013 .

[22]  J. Verwer,et al.  Numerical solution of time-dependent advection-diffusion-reaction equations , 2003 .

[23]  Chad Lieberman,et al.  On Adaptive Choice of Shifts in Rational Krylov Subspace Reduction of Evolutionary Problems , 2010, SIAM J. Sci. Comput..

[24]  Xiaoye S. Li,et al.  An overview of SuperLU: Algorithms, implementation, and user interface , 2003, TOMS.

[25]  Luca Antiga,et al.  Automatic differentiation in PyTorch , 2017 .

[26]  Liliana Borcea,et al.  Robust nonlinear processing of active array data in inverse scattering via truncated reduced order models , 2018, J. Comput. Phys..

[27]  John L. Nazareth,et al.  Introduction to derivative-free optimization , 2010, Math. Comput..

[28]  Yousef Saad,et al.  Iterative methods for sparse linear systems , 2003 .

[29]  Vladimir Druskin,et al.  On Optimal Convergence Rate of the Rational Krylov Subspace Reduction for Electromagnetic Problems in Unbounded Domains , 2009, SIAM J. Numer. Anal..

[30]  V. Lebedev,et al.  Solving reactor problems to determine the multiplication: A new method of accelerating outer iterations , 2007 .

[31]  Vladimir Druskin,et al.  Solution of Large Scale Evolutionary Problems Using Rational Krylov Subspaces with Optimized Shifts , 2009, SIAM J. Sci. Comput..

[32]  C. Sierra,et al.  Linear Autonomous Compartmental Models as Continuous-Time Markov Chains: Transit-Time and Age Distributions , 2017, Mathematical Geosciences.