Sigmoidal Weight Constraint in a Recurrent Neural Network

Abstract

When training a recurrently connected neural network (RNN), the magnitude of the connection strengths (weights) must be limited in some way. The weights are normally constrained by either renormalizing them after each learning step, or by using a decay term proportional to the weight. For large numbers of training cycles, we show that an RNN output can become unstable with previously used weight adjustment methods. We introduce a technique that constrains weight values to move on a smooth sigmoidal curve. Without the need for renormalization or a parametric decay term, our RNNs then produce stable output. Performance is also improved in other ways. As an example, an associative memory RNN is shown to converge much faster and to more accurate values than with previous methods.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 2004
Accession Number
ADA436300

Entities

People

  • Simon A. Barton

Tags

Communities of Interest

  • Materials and Manufacturing Processes
  • Sensors

DTIC Thesaurus Topics

  • Abstracts
  • Accuracy
  • Algorithms
  • Artificial Intelligence
  • Artificial Intelligence Computing
  • Artificial Intelligence Software
  • Autonomous Vehicles
  • Classification
  • Computers
  • Content Addressable Memory
  • Identification
  • Learning
  • Neural Networks
  • Pattern Recognition
  • Recurrent Neural Networks
  • Security
  • Training

Readers

  • Calculus or Mathematical Analysis
  • Small Business Innovation Research Program (SBIR) EDI Research and Innovation.
  • Theoretical Analysis.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms
  • AI & ML - Neural Networks