Analytical derivatives of neural networks

DOI10.1016/J.CPC.2021.108169MaRDI QIDQ6162825zbMATH OpenOpenAlexFDO

Authors Simone Rodini

Publication date 16 June 2023

Published in Computer Physics Communications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2105.02737

zbMATH Keywords

deep neural networksPINNvariational neural network

Mathematics Subject Classification ID

Artificial intelligence (68Txx)

Abstract: We propose a simple recursive algorithm that allows the computation of the first- and second-order derivatives with respect to the inputs of an arbitrary deep feed forward neural network (DFNN). The algorithm naturally incorporates the derivatives with respect to the network parameters. To test the algorithm, we apply it to the study of the quantum mechanical variational problem for few cases of simple potentials, modeling the ground-state wave function in terms of a DFNN.

Cites work

• Approximation by superpositions of a sigmoidal function
• DGM: a deep learning algorithm for solving partial differential equations
• Modern Quantum Mechanics
• Neural algorithm for solving differential equations
• Numerical solution of the Schrödinger equation by neural network and genetic algorithm
• The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems

Cited in

(1)

• Physics-augmented neural networks for constitutive modeling of hyperelastic geometrically exact beams

This page was built for publication: Analytical derivatives of neural networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6162825)