Algorithm/Architecture Study for Artificial Neural Nets

Abstract

Neural information processing has already helped catalyzed many potential opportunities of cross-fertilization, from which many diversified disciplines have benefited mutually. However, in order to substain a long-term impact, there must establish a fundamental and coherent theory for it. This project focuses on the development and understanding of the fundamental system theoretical basis for temporal dynamic networks. The main thrust of the research hinges on a thorough understanding of several key issues regarding temporal dynamical system modeling, including model unification, training efficiency, generalization performance, and hierarchical network structure.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 30, 1993
Accession Number
ADA271820

Entities

People

  • Sun Yuan Kung

Organizations

  • Princeton University

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Abstracts
  • Algorithms
  • Character Recognition
  • Computational Science
  • Computing System Architectures
  • Hidden Markov Models
  • Information Processing
  • Information Science
  • Markov Models
  • Neural Networks
  • Numerical Analysis
  • Pattern Recognition
  • Recognition
  • Recurrent Neural Networks
  • Signal Processing
  • Simulators
  • Training

Readers

  • Integrated Circuit Design and Technology.
  • Neural Network Machine Learning.
  • Systems Analysis and Design