DeepAI AI Chat
Log In Sign Up

Generalization of partitioned Runge–Kutta methods for adjoint systems

by   Takeru Matsuda, et al.

This study computes the gradient of a function of numerical solutions of ordinary differential equations (ODEs) with respect to the initial condition. The adjoint method computes the gradient approximately by solving the corresponding adjoint system numerically. In this context, Sanz-Serna [SIAM Rev., 58 (2016), pp. 3–33] showed that when the initial value problem is solved by a Runge–Kutta (RK) method, the gradient can be exactly computed by applying an appropriate RK method to the adjoint system. Focusing on the case where the initial value problem is solved by a partitioned RK (PRK) method, this paper presents a numerical method, which can be seen as a generalization of PRK methods, for the adjoint system that gives the exact gradient.


page 1

page 2

page 3

page 4


Numerical Optimal Control of HIV Transmission in Octave/MATLAB

We provide easy and readable GNU Octave/MATLAB code for the simulation o...

Adjoint-based exact Hessian-vector multiplication using symplectic Runge–Kutta methods

We consider a function of the numerical solution of an initial value pro...

Comparison of Numerical Solvers for Differential Equations for Holonomic Gradient Method in Statistics

Definite integrals with parameters of holonomic functions satisfy holono...

Fenrir: Physics-Enhanced Regression for Initial Value Problems

We show how probabilistic numerics can be used to convert an initial val...

Accelerating Neural ODEs with Spectral Elements

This paper proposes the use of spectral element methods canuto_spectral_...

Extraction of a computer-certified ODE solver

Reliably determining system trajectories is essential in many analysis a...

ODE Test Problems: a MATLAB suite of initial value problems

ODE Test Problems (OTP) is an object-oriented MATLAB package offering a ...