Abstract
This paper compares Neurolinear feature extraction technique to two other feature extraction techniques. The author has developed all three methods recently, but in this paper, Neurolinear method has been improved even further. The study is performed in respect to performance, accuracy and suitability for visualisation. Experimental study is performed on a variety of datasets from numerous domains: Chemistry, biology, finance etc. In total, 19 real and two artificial datasets were used. Comparison is performed in terms of computational efficiency and accuracy. Results show that Neurolinear method achieves best accuracy while Best directions method is not so accurate but often much faster.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Raudys, A.: A nonparametric data mapping technique for active initialization of the multilayer perceptron. In: Amin, A., et al. (eds.) Advances in Pattern Recognition, pp. 989–996. Springer, Berlin Heidelberg New York (1998)
Raudys, A., Long, J.A.: MLP based linear feature extraction for nonlinearly separable data. Pattern Analysis & Applications 4(4), 227–234 (2001)
Raudys, A.: High speed associative memories for feature extraction and visualisation. Pattern Recogn. Lett. 24(9/10), 1327–1339 (2003)
Haykin, S.: In: Edition, S. (ed.) Neural Networks: A Comprehensive Foundation. Macmillan College, New York (1999)
Sammon, J.W.: A nonlinear mapping for data structure analysis. IEEE Trans. Comput. 18, 401–409 (1969)
Foley, D.H., Sammon, J.W.J.: An optimal set of discriminant vectors. IEEE Trans. Comput. 24, 281–289 (1975)
Raudys, A.: Boosting neural network feature extraction by reduced accuracy activation functions. Pattern Recogn. 36(6), 1343–1354 (2003)
Raudys, A.: A Non-parametric data mapping technique for active initialisation of the multilayer perceptron. In: Joint IAPR Int. Workshops/SSPR'98 and SPR'98. Springer, Berlin Heidelberg New York (1998)
Lloyd, S.P.: Least squares quantisation in PCM. IEEE Trans. Inf. Theory 28(2), 129–137 (1982)
Duda, R.O., Hart, P.E., Stork, D.G.: Pattern Classification, 2nd ed. Wiley, New York (2000)
Light, W. (ed.): Advances in Numerical Analysis: Wavelets, Subdivision Algorithms, and Radial Basis Functions. Clarendon, Oxford (1992)
Austin, J.: ADAM: A distributed associative memory for scene analysis. In: First International Conference on Neural Networks. San Diego (1987)
Austin, J.: Distributed associative memories for high-speed symbolic reasoning. Int. J. Fuzzy Sets Syst. 82, 223–233 (1995)
Austin, J.: An associative neural architecture for invariant pattern classification. In: IEE International Conference on Artificial Neural Networks. IEE Conf. (1989)
Austin, J.: High speed invariant pattern recognition using adaptive neural networks. In: IEE Proceedings of Conference on Image Processing and its Applications. IEE Conference Publication, University of Warwick, UK (1989)
Austin, J., Lees, K., Kennedy, J.: Reasoning with Neural Networks. In IEE Colloquium “Real Time Knowledge Based Systems” (1995)
Austin, J., Lees, K.: A search engine based on neural correlation matrix memories. Neurocomputing 35(1–4): 55–72 (2000)
Davis, L.: Handbook of Genetic Algorithms. Van Nostrand Reinhold (1991)
Raudys, A., Long, J.A.: Modelling company credit ratings using a number of classification techniques. In: Conference on Cybernetics and Systems Research. Austrian Society for Cybernetic Studies, Vienna (2000)
Saudargiene, A.: Structurization of the covariance matrix by process type and block diagonal models in the classifier design. Informatica 10(2), 245–269 (1999)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Raudys, A. Comparative Study of Improved Neurolinear Method to Two Other Novel Feature Extraction Techniques. J Intell Robot Syst 46, 163–180 (2006). https://doi.org/10.1007/s10846-006-9058-0
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10846-006-9058-0