Kernel-based immunity synergetic network for image classification

ABSTRACT

In order to reduce the relativity among prototype pattern vectors and to enhance the separability among different patterns, a novel kernel-based learning algorithm of Synergetic Neural Network (SNN) is proposed. The algorithm first maps the data from original space into a new feature space and then classifies them by a two-layered SNN. An optimization method of weighted factors in the two-layered SNN is also presented. It gives different patterns to different weights and makes full use of the global and local searching ability of Immunity Clonal Algorithm (ICA). Experiments on Iris dataset, textural images and Synthetic Aperture Radar (SAR) images show that the new algorithm does not only improve the classification rate but also has shorter training and testing time.

REFERENCES

• 1: Hermann Haken, Synergetic computers and cognition: a top-down approach to neural nets, Springer-Verlag New York, Inc., New York, NY, 1991
• 2: Chen, W. G. and Qi, F. H. 2004. A novel learning algorithm of synergetic pattern recognition. Journal of Shanghai Jiaotong University. Vol.38, No.1, pp.18--20.
• 3: Ren, Q. S., Qiao, H., and Chen, B. 2003. Image recognition using a quadratic convergent learning algorithm of synergetic neural network. In Proceedings the 2003 IEEE International Conference on Robotics, Intelligent Systems and Signal Processing (Changsha, China, October, 2003). pp.255--259.
• 4: Jiao, L. C. and Du, H. F. 2003. Development and prospect of the artificial immune system. Acta Electronica Sinica. Vol.31, No.10, pp.1540--1548.
• 5: T. Wagner , F. G. Boebel, Testing synergetic algorithms with industrial classification problems, Neural Networks, v.7 n.8, p.1313-1321, 1994 [doi>10.1016/0893-6080(94)90012-4]
• 6: Ma, X. L. and Jiao L. C. 2004. An effective learning algorithm of synergetic neural network. Lecture Notes in Computer Science. Vol.3173, pp.258--263.
• 7: Ma, X. L. and Jiao, L. C. 2004. Reconstruction of order parameters based on immunity clonal strategy for image classification. Lecture Notes in Computer Science. Vol.3211, pp.455--462.
• 8: Meyer, F. G. and Coifman, R. R. 1997. Brushlets: a tool for directional image analysis and image compression. Applied and Computational Harmonic Analysis. Vol.4, No.2, pp. 147--187.

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.5 PATTERN RECOGNITION
      I.5.1 Models
          Subjects: Neural nets

General Terms:
Algorithms, Experimentation

Keywords:
image classification., immunity clonal algorithm, kernel learning, synergetic neural network