Applied Technologies and Innovations

  Previous Article | Back to Volume | Next Article
  Abstract | References | Citation | Download | Preview | Statistics
Volume 4
Issue 1
Online publication date 2011-04-18
Title Methods and algorithms of adaptive designing for neuronetworking system of processing the data with non-stationary nature
Author Olimjan Djumanov
Abstract The hybrid model of data neuronetworking processing system, in which the opportunities of a wide spectrum of methods and algorithms of neural network and statistical models are combined, is researched. The original ways and algorithms of adaptation during designing of neural network are developed by escalating, selection and adjustment of activation functions. The strategy and new principles are developed for neural network output quality control. The theoretical results were approved on the basis of Koxonen model; and as result conditions of well over reduction of output data approximation error are proved.
Citation
References
Banarse, D., 1997. “A generic neural network architecture for deformation invariant object recognition,” Unpublished PhD Thesis, University of Wales, Bangor, pp.56-64. 

Basit Hussain and Kabuka, M., 1994. “A novel feature recognition neural network and its application to character recognition,” IEEE Trans. on Pattern Analysis and Machine Intelligence, 16(1), p.98-106. 

Boucher, A., Hidalgo, P., Thonnat, M., Belmonte, J., Galan, C., Bonton, P., and Tomczak, R., 2002. “Development of a semi-automatic system for pollen recognition,” Aerobiologia, 18(3-4), pp.195-20.

Djumanov, O.,  2007. “Programmed system for the adaptive monitoring of the continuous nature information on the basis of supervised learning of neural network,” IT Promotion in Asia, TUIT, Tashkent, Information Technology Internationalization Research Center, pp.181-90.
 
Djumanov, O., 2008. “Adaptive training of intellectual system of processing and analysis of the non-stationary on nature information,” Fifth World Conference on Intelligent Systems for Industrial Automation, 25-27 November, Tashkent, Uzbekistan, pp.133-40.

Djumanov, O., 2009. “The adaptive control of accuracy at transfer and processing of the continuous information in view of non-stationary process,” IT Promotion in Asia, Tashkent University of IT, Uzbekistan, ITIRC, Korea, pp.116-20.

Kohonen, T., 1990. “Improved versions of learning vector quantization,” Proceedings of the International Joint Conference on Neural Networks, San Diego: IEEE, Vol.I, pp.545-50.

Widrow, B. and Lehr, A., 1990. “30 years of adaptive neural networks: Perception, madaline and Backpropagation,” In: Proc. IEEE 75 (1990), 9, 1415-442.

Keywords Architecture of neural network, adaptation, nonstationarity, activation function, recognition of micro objects.
DOI http://dx.doi.org/10.15208/ati.2011.6
Pages 48-57
Download Full PDF Download
  Previous Article | Back to Volume | Next Article
Share
Search in articles
Statistics
Journal Published articles
ATI 263
Journal Hits
ATI 789669
Journal Downloads
ATI 7698
Total users online -