Models for Modular Neural Networks: A Comparison Study

Eva Volna

2009

Abstract

There are mainly two approaches for machine learning: symbolic and sub-symbolic. Decision tree is a typical model for symbolic learning, and neural network is a model for sub-symbolic learning. For pattern recognition, decision trees are more efficient than neural networks for two reasons. First, the compu-tations in making decisions are simpler. Second, important features can be se-lected automatically during the design process. This paper introduces models for modular neural network that are a neural network tree where each node being an expert neural network and modular neural architecture where interconnections between modules are reduced. In this paper, we will study adaptation processes of neural network trees, modular neural network and conventional neural network. Then, we will compare all these adaptation processes during experimental work with the Fisher's Iris data set that is the bench test database from the area of machine learning. Experimental results with a recognition problem show that both models (e.g. neural network tree and modular neural network) have better adaptation results than conventional multilayer neural network architecture but the time complexity for trained neural network trees increases exponentially with the number of inputs.

References

  1. Takeda, T., Zhao Q. F., and Liu, Y.: A study on on-line learning of NNTrees. In: Proc. International Joint Conference on Neural Networks, pp. 145-152, (2003).
  2. Zhao, Q. F.: Evolutionary design of neural network tree -integration of decision tree, neural network and GA. In: Proc. IEEE Congress on Evolutionary Computation, pp. 240-244, Seoul, (2001).
  3. Mizuno, S., and Zhao, Q. F.: Neural network trees with nodes of limited inputs are good for learning and understanding. In: Proc. 4th Asia-Pacific Conference on Simulated Evolution And Learning, pp. 573-576, Singapore, (2002).
  4. Endou, T. and Zhao, Q.F.: Generation of comprehensible decision trees through evolution of training data. In Proc. IEEE Congress on Evolutionary Computation (CEC'2002) Hawai, 2002.
  5. Zhao, Q. F.: Modelling and evolutionary learning of modular neural networks. In: Proc. The 6-th International symposiums on artificial life and robotics, pp.508-511, Tokyo (2001).
  6. Quinlan, J. R.: C4.5: Programs for machine learning. Morgan Kaufmann Publishers, (1993).
  7. http://en.wikipedia.org/wiki/Iris_flower_data_set (from 16/1/2009).
  8. Di Fernando, A., Calebretta, R., and Parisi, D.: Evolving modular architectures for neural networks. In French R., and Sougne, J. (eds.).Proceedings of the Sixth Neural Computation and Psychology Workshop: Evolution, Learning and Development. Springer Verlag, London, 2001.
  9. Fausett, V. L.: Fundamental of Neural Networks, Architecture, Algorithms and Applications, Prentice Hall; US Ed edition, 1994.
Download


Paper Citation


in Harvard Style

Volna E. (2009). Models for Modular Neural Networks: A Comparison Study . In Proceedings of the 5th International Workshop on Artificial Neural Networks and Intelligent Information Processing - Volume 1: Workshop ANNIIP, (ICINCO 2009) ISBN 978-989-674-002-3, pages 23-30. DOI: 10.5220/0002196700230030


in Bibtex Style

@conference{workshop anniip09,
author={Eva Volna},
title={Models for Modular Neural Networks: A Comparison Study},
booktitle={Proceedings of the 5th International Workshop on Artificial Neural Networks and Intelligent Information Processing - Volume 1: Workshop ANNIIP, (ICINCO 2009)},
year={2009},
pages={23-30},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0002196700230030},
isbn={978-989-674-002-3},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 5th International Workshop on Artificial Neural Networks and Intelligent Information Processing - Volume 1: Workshop ANNIIP, (ICINCO 2009)
TI - Models for Modular Neural Networks: A Comparison Study
SN - 978-989-674-002-3
AU - Volna E.
PY - 2009
SP - 23
EP - 30
DO - 10.5220/0002196700230030