Skip to main content
Presentation
Investigation of Weight Reuse in Multi-Layer Perceptron Networks forAccelerating the Solution of Differential Equations
Proceedings of the 4th International Conference on Intelligent SystemsDesign and Applications (2004)
  • Kevin McFall, Kennesaw State University
  • J. R. Mahan
Abstract
Research has shown that training multi-layer perceptron networks to solve ordinary and partial differential
equations (DEs) can be accelerated by reusing network weights from a previously solved similar problem. This paper compares weight reuse for two existing methods of defining the network error function. Weight reuse is shown to accelerate training of one ordinary and two partial DEs even for equations with significantly different parameters or boundary/initial conditions. The second method outperforms the first for partial DEs where multiple boundary/initial conditions are defined, but fails unpredictably when weight reuse is applied to accelerate solution of the diffusion equation.
Disciplines
Publication Date
August, 2004
Citation Information
Kevin McFall and J. R. Mahan. "Investigation of Weight Reuse in Multi-Layer Perceptron Networks forAccelerating the Solution of Differential Equations" Proceedings of the 4th International Conference on Intelligent SystemsDesign and Applications (2004)
Available at: http://works.bepress.com/kevin-mcfall/23/