Bayesian numerical methods for nonlinear partial differential equations
نویسندگان
چکیده
The numerical solution of differential equations can be formulated as an inference problem to which formal statistical approaches applied. However, nonlinear partial (PDEs) pose substantial challenges from inferential perspective, most notably the absence explicit conditioning formula. This paper extends earlier work on linear PDEs a general class initial value problems specified by PDEs, motivated for evaluations right-hand-side, conditions, or boundary conditions PDE have high computational cost. proposed method viewed exact Bayesian under approximate likelihood, is based discretisation operator. Proof-of-concept experimental results demonstrate that meaningful probabilistic uncertainty quantification unknown performed, while controlling number times and are evaluated. A suitable prior model identified using novel theoretical analysis sample path properties Mat\'{e}rn processes, may independent interest.
منابع مشابه
global results on some nonlinear partial differential equations for direct and inverse problems
در این رساله به بررسی رفتار جواب های رده ای از معادلات دیفرانسیل با مشتقات جزیی در دامنه های کراندار می پردازیم . این معادلات به فرم نیم-خطی و غیر خطی برای مسایل مستقیم و معکوس مورد مطالعه قرار می گیرند . به ویژه، تاثیر شرایط مختلف فیزیکی را در مساله، نظیر وجود موانع و منابع، پراکندگی و چسبندگی در معادلات موج و گرما بررسی می کنیم و به دنبال شرایطی می گردیم که متضمن وجود سراسری یا عدم وجود سراسر...
A Review of Numerical Methods for Nonlinear Partial Differential Equations
Numerical methods were first put into use as an effective tool for solving partial differential equations (PDEs) by John von Neumann in the mid1940s. In a 1949 letter von Neumann wrote “the entire computing machine is merely one component of a greater whole, namely, of the unity formed by the computing machine, the mathematical problems that go with it, and the type of planning which is called ...
متن کاملRobust Numerical Methods for Partial Differential Equations
The general theme of this project is to study numerical methods for systems of partial diffential equations which depend on one or more critical parameters. Typically, we are interested in systems which change type as a critical perturbation parameter tend to zero. Our goal is to construct numerical methods with convergence properties which are uniform with respect to the perturbation parameter...
متن کاملNumerical Methods for Fuzzy Linear Partial Differential Equations under new Definition for Derivative
In this paper difference methods to solve "fuzzy partial differential equations" (FPDE) such as fuzzy hyperbolic and fuzzy parabolic equations are considered. The existence of the solution and stability of the method are examined in detail. Finally examples are presented to show that the Hausdorff distance between the exact solution and approximate solution tends to zero.
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Statistics and Computing
سال: 2021
ISSN: ['0960-3174', '1573-1375']
DOI: https://doi.org/10.1007/s11222-021-10030-w