Stochastic rounding and reduced-precision fixed-point arithmetic for solving neural ordinary differential equations
نویسندگان
چکیده
منابع مشابه
Artificial neural networks for solving ordinary and partial differential equations
We present a method to solve initial and boundary value problems using artificial neural networks. A trial solution of the differential equation is written as a sum of two parts. The first part satisfies the initial/boundary conditions and contains no adjustable parameters. The second part is constructed so as not to affect the initial/boundary conditions. This part involves a feedforward neura...
متن کاملFixed point theorems for $alpha$-$psi$-contractive mappings in partially ordered sets and application to ordinary differential equations
In this paper, we introduce $alpha$-$psi$-contractive mapping in partially ordered sets and construct fixed point theorems to solve a first-order ordinary differential equation by existence of its lower solution.
متن کاملCirculant Preconditioners for Solving Ordinary Differential Equations
In this paper, we consider the solution of ordinary diierential equations using boundary value methods. These methods require the solutions of one or more unsymmetric, large and sparse linear systems. Krylov subspace methods with the Strang block-circulant preconditioners are proposed for solving these linear systems. We prove that our preconditioners are invertible and all the eigenvalues of t...
متن کاملInvariant functions for solving multiplicative discrete and continuous ordinary differential equations
In this paper, at first the elemantary and basic concepts of multiplicative discrete and continous differentian and integration introduced. Then for these kinds of differentiation invariant functions the general solution of discrete and continous multiplicative differential equations will be given. Finaly a vast class of difference equations with variable coefficients and nonlinear difference e...
متن کاملAn Interpretation of Stochastic Differential Equations as Ordinary Differential Equations Which Depend on the Sample Point
where ww(/) = Wt(cS). Usually, this approach is not followed because of the technical difficulties that appear when one tries to solve (det) for arbitrary continuous w. One is therefore forced to use other lines of attack, and to study (stoch) directly. This gives rise to at least two nonequivalent theories, namely, the one due to Ito, and that of Fisk and Stratonovich. The purpose of this note...
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: Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences
سال: 2020
ISSN: 1364-503X,1471-2962
DOI: 10.1098/rsta.2019.0052