Examining a Hybrid Connectionist/Symbolic System for the Analysis of Ballistic Signals

  • Charles Lin
  • James Hendler
Part of the The Springer International Series In Engineering and Computer Science book series (SECS, volume 292)


The field of artificial intelligence (AI) has produced a variety of different problem solving paradigms. Two of the more prominent ones are symbolic AI and connectionism. Some researchers [4] [14] have argued that symbolism and connectionism represent differing computational paradigms, while others have discussed merging the differing strengths and weaknesses of these approaches, as evidenced by the papers in this volume. The general consensus to date seems to be been that symbolic systems are currently better at reasoning tasks and encapsulating expert knowledge, while connectionist systems have been used more successfully in pattern recognition and other perceptual tasks. Interestingly enough, the strengths of symbolic systems correspond to the weaknesses of connectionist systems and vice versa.


Neural Network Expert System Hybrid System Input Vector Symbolic System 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    I. Bruja (1992). Neural Net Approach to Recognition of Waveforms, Artificial Intelligence and Tutoring Systems for Teaching and Learning, pp. 43–63, London: Ellis Horwood Limited.Google Scholar
  2. [2]
    Y. Cui, J. Hendler, H. Su, and T. McAvoy. (in press). Improving PID controllers with a Neural Network. Tech. rep., Institute of Systems Research, University of Maryland, 1993.Google Scholar
  3. [3]
    J. Dunker, A. Scherer, and G. Schlageter. (1992). Integrating Neural Networks into a Distributed Knowledge Base. In Int. Conf. on AL Expert Systems and Neural Language.Google Scholar
  4. [4]
    M. Dyer. (1988). Symbolic NeuroEngineering for Natural Language Processing: A Multilevel Research Approach. Tech. Rep. UCLA-AI-88-14, Computer Science Dept., University of California, Los Angeles.Google Scholar
  5. [5]
    G.M. Scott, J.W. Shavlik and W.H. Ray. (1991). Refining PID Controllers using Neural Networks, In J. Moody, S. Hanson and R. Lippmann (Eds.), Advances in Neural Information Processing Systems 4, pp. 555–562, Morgan Kaufmann: San Mateo, CA.Google Scholar
  6. [6]
    R. M. Goodman, C. M. Higgins, J. W. Miller, and P. Smyth. (1992). Rule-Based Neural Networks for Classification and Probability Estimation. Neural Computation, No. 4, pp. 781–804.CrossRefGoogle Scholar
  7. [7]
    R. M. Goodman, J. W. Miller, and P. Smyth (1989). An Information Theoretic Approach to Rule-Based Connectionist Expert System, Advances in Neural Information Processing Systems 7, pp. 256–263, San Mateo, CA: Morgan Kaufmann.Google Scholar
  8. [8]
    H. K. Greenspan, R. Goodman, and R. Chellappa. (1992). Combined Neural Network and Rule-Based Framework for Probabilistic Pattern Recognition and Discovery, Advances in Neural Information Processing Systems 4, pp. 444–451, San Mateo, CA: Morgan Kaufmann.Google Scholar
  9. [9]
    D. A. Handelman, S. H. Lane, and J. J. Gelfand. (1990). Integrating Neural Networks for Intelligent Robot Control. IEEE Control System Magazine, pp. 77–87.Google Scholar
  10. [10]
    J. Hendler and L. Dickens. (1992). Integrating Neural Network and Expert Reasoning: An Example. Proceedings ofAISB, pp. 109–116.Google Scholar
  11. [11]
    L. Kanal and S. Raghavan. (1992). Hybrid Systems-A Key to Intelligent Pattern Recognition. International Joint Conference of Neural Networks, Vol. IV of IV, June 7–11.Google Scholar
  12. [12]
    S. G. Romaniuk and L. O. Hall. (1993). SC-net: A Hybrid Connectionist, Symbolic System. Information Sciences.Google Scholar
  13. [13]
    T. Sejnowski and C. R. Rosenberg. (1986). NETtalk: a parallel network that learns to read aloud. Tech. Rep. EECS-86/01, Johns Hopkins Univ.Google Scholar
  14. [14]
    P. Smolensky. (1990). Connectionist AI, Symbolic AI, and the Brain. AI Review.Google Scholar
  15. [15]
    G. G. Towell, J. W. Shavlik, and M. O. Noordweier. (1990). Refinement of Approximate Domain Theories by Knowledge-Based Neural Networks. In Proceedings, AAAI-90, pp. 861–866.Google Scholar
  16. [16]
    M. E. Ulug. (1989). A Hybrid Expert System Combining AI Techniques with a Neural-Net. In Proceedings of the Second International Conference on Industrial and Engineering Applications of AI and Expert Systems, pp. 305–309.Google Scholar

Copyright information

© Kluwer Academic Publishers 1995

Authors and Affiliations

  • Charles Lin
    • 1
  • James Hendler
    • 1
  1. 1.Department of Computer ScienceUniversity of MarylandCollege Park

Personalised recommendations