Extended self organizing feature map: A tagged potential field approach

Date

1999-8

Author

Baykal, Nazife

Metadata

Show full item record

Item Usage Stats

117
views

0
downloads

This paper proposes an escape methodology to the local minima problem of self organizing feature maps generated in the overlapping regions which are equidistant to the corresponding winners. Two new versions of the Self Organizing Feature Map are derived equipped with such a methodology. The first approach introduces an excitation term, which increases the convergence speed and efficiency of the algorithm, while increasing the probability of escaping from local minima. In the second approach, we associate a learning set which specifies the attractive and repulsive field of output neurons. Results indicate that accuracy percentile of the new methods are higher than the original algorithm while they have the ability to escape from local minima.

Subject Keywords

Self Organizing Feature Map, Neural Modeling, Local Minima Avoidance, Classification, Radial Basis Function Networks

URI

https://hdl.handle.net/11511/52496

Journal

NEURAL PROCESSING LETTERS

DOI

https://doi.org/10.1023/A:1018662813803

Collections

Graduate School of Informatics, Article

Suggestions

OpenMETU
Core

Tagged potential field extension to self organizing feature maps Baykal, Nazife; Erkmen, Aydan Müşerref (1998-04-23) This paper proposes an escape methodology to the local minima problem of self organizing feature maps generated in the overlapping regions which are equidistant to the corresponding winners. Two new versions of self organizing feature map are derived equipped with such a methodology. The first approach introduces an excitation term, which increases the convergence speed and efficiency of the algorithm while increasing the probability of escaping from local minima. In the second approach we associate a learn...
A novel music algorithm based electromagnetic target recognition method in resonance region for the classification of single and multiple targets Seçmen, Mustafa; Sayan, Gönül; Department of Electrical and Electronics Engineering (2008) This thesis presents a novel aspect and polarization invariant electromagnetic target recognition technique in resonance region based on use of MUSIC algorithm for the extraction of natural-resonance related target features. In the suggested method, the feature patterns called “MUSIC Spectrum Matrices (MSMs)” are constructed for each candidate target at each reference aspect angle using targets’ scattered data at different late-time intervals. These individual MSMs correspond to maps of targets’ natural-res...
A probabilistic multiple criteria sorting approach based on distance functions ÇELİK, BİLGE; Karasakal, Esra; İyigün, Cem (2015-05-01) In this paper, a new probabilistic distance based sorting (PDIS) method is developed for multiple criteria sorting problems. The distance to the ideal point is used as a criteria disaggregation function to determine the values of alternatives. These values are used to sort alternatives into the predefined classes. The method also calculates probabilities that each alternative belong to the predefined classes in order to handle alternative optimal solutions. It is applied to five data sets and its performanc...
A framework for state attraction of discrete event systems under partial observation Schmidt, Klaus Verner (2014-10-10) State attraction for discrete event systems (DES) addresses the problem of reaching a desired subset of the plant state space after a bounded number of event occurrences. The problem of state attraction arises for example in fault-tolerant supervisory control or in the control of reconfigurable manufacturing systems, and is also applicable to systems biological problems such as the control of gene regulatory networks. State attraction is investigated with the assumption of full event observation in the exi...
A Radar Target Recognition Method with MUSIC Algorithm: Application to Aircraft Targets with Measured Scattered Data Secmen, M.; Turhan-Sayan, G.; Sayan, Gönül (2008-05-30) This paper demonstrates the usefulness of an ultra wideband target recognition method in the case of realistic and complicated target geometries at resonance region. The method utilizes the MUSIC algorithm to extract the natural resonance-related scattering features of targets. The resulting features give the power distribution maps of targets. These maps are called as fused MUSIC spectrum matrices and used as the main target recognition feature in the method. The fusion process is needed to reduce the aspe...

Citation Formats

N. Baykal, “Extended self organizing feature map: A tagged potential field approach,” NEURAL PROCESSING LETTERS, pp. 57–72, 1999, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/52496.