Adjoint Operator Algorithms for Faster Learning in Dynamical Neural Networks

J. Barhen, Nikzad Toomarian, Sandeep Gulati

Neural Information Processing Systems · 1989 · 14 citations · 9 references

Concepts

Abstract

A methodology for faster supervised learning in dynamical nonlinear neural networks is presented. It exploits the concept of adjoint operators to enable computation of changes in the network's response due to perturbations in all system parameters, using the solution of a single set of appropriately constructed linear equations. The lower bound on speedup per learning iteration over conventional methods for calculating the neuromorphic energy gradient is O(N2), where N is the number of neurons in the network.

References

9