In this paper, we propose a new feature extraction method for feedforward neural networks. The method is based on the recently published decision boundary feature extraction algorithm which is based on the fact that all the necessary features for classification can be extracted from the decision boundary. The decision boundary feature extraction algorithm can take advantage of characteristics of neural networks which can solve complex problems with arbitrary decision boundaries without assuming underlying probability distribution functions of the data. To apply the decision boundary feature extraction method, we first give a specific definition for the decision boundary in a neural network. Then, we propose a procedure for extracting all the necessary features for classification from the decision boundary. Experiments show promising results.
Bibliographical noteFunding Information:
Manuscript received December 5, 1995; revised June 20, 1996. This work was supported in part by NASA under Grant NAGW-925. C. Lee was with the School of Electrical and Computer Engineering, Purdue University, W. Lafayette, IN 47907-1285 USA. He is now with the Department of Electrical Engineering, Yonsei University, Seoul, Korea. D. A. Landgrebe is with the School of Electrical and Computer Engineering, Purdue University, W. Lafayette, IN 47907-1285 USA. Publisher Item Identifier S 1045-9227(97)00236-1.
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- Computer Networks and Communications
- Artificial Intelligence