An Efficient Extraction of Data Feature By Using Neural Networks of Hybrid Learning Algorithm


The KIPS Transactions:PartB , Vol. 8, No. 2, pp. 130-136, Apr. 2001
10.3745/KIPSTB.2001.8.2.130,   PDF Download:

Abstract

This paper proposes an efficient feature extraction method of the image data using nonlinear principal component analysis (NLPCA) neural networks of a new learning algorithm. The proposed learning algorithm is a hybrid algorithm combined momentum and dynamic tunneling. The momentum is applied for high-speed convergence by restraining an oscillation in the process of converging to the optimal solution. Converging to the local minimum, the dynamic tunneling is also applied for estimating the new initial weights for converging to the global minimum. The proposed algorithm has been applied to a cancer image of 256X256 pixels and a face image of 128X128 pixels, respectively. The experiment results show that the proposed algorithm has better performances of the convergence and the nonlinear feature extraction than those using the conventional NLPCA neural networks.


Statistics
Show / Hide Statistics

Statistics (Cumulative Counts from September 1st, 2017)
Multiple requests among the same browser session are counted as one view.
If you mouse over a chart, the values of data points will be shown.


Cite this article
[IEEE Style]
Y. H. Cho, J. H. Yun, Y. S. Park, "An Efficient Extraction of Data Feature By Using Neural Networks of Hybrid Learning Algorithm," The KIPS Transactions:PartB , vol. 8, no. 2, pp. 130-136, 2001. DOI: 10.3745/KIPSTB.2001.8.2.130.

[ACM Style]
Yong Hyun Cho, Jung Hwan Yun, and Yong Soo Park. 2001. An Efficient Extraction of Data Feature By Using Neural Networks of Hybrid Learning Algorithm. The KIPS Transactions:PartB , 8, 2, (2001), 130-136. DOI: 10.3745/KIPSTB.2001.8.2.130.