×
Oh, Sung-Kwun; Kim, Dong-Won; Pedrycz, Witold

Hybrid fuzzy polynomial neural networks. (English) Zbl 1075.93020

Int. J. Uncertain. Fuzziness Knowl.-Based Syst. 10, No. 3, 257-280 (2002).
The authors study a new neurofuzzy topology, called Hybrid Fuzzy Polynomial Neural Networks (HFPNN), and discuss a comprehensive design methodology supporting their development. HFPNN is a network resulting from the combination of a fuzzy inference system and a Polynomial Neural Network (PNN) algorithm. Each node of the first layer of the HFPNN that is a fuzzy polynomial neuron (FPN) operates as a compact fuzzy inference system. By exploiting several types of regression polynomials in the conclusion part of the rules, the architecture of the HFPNN can be easily changed to adapt to the system environment. The networks of the second and higher layers of the HFPNN come with a high level of flexibility, as each node (processing element forming a partial description (PD)) can exploit a different order of the polynomial (say, linear, quadratic, cubic, etc.). HFPNN is a flexible neural architecture whose structure is based on the Group Method of Data Handling (GMDG) and developed through learning. In particular, the number of layers of the PNN is not fixed in advance but is generated in a dynamic way. The experimental part of the study involves two representative numerical examples such as a chaotic time series and Box-Jenkins gas furnace data.
Reviewer: D. Volný (Praha)

Cited in 3 Documents

MSC:

93C42 Fuzzy control/observation systems
92B20 Neural networks for/in biological studies, artificial life and related topics
68T05 Learning and adaptive systems in artificial intelligence

Keywords:

hybrid architecture; fuzzy rule-based computing; polynomial neural networks (PNN); group method of data handling (GMDH); fuzzy polynomial neuron (FPN); time series data
Cite Review PDF
Full Text: DOI

References:

[1] DOI: 10.1016/0893-6080(89)90020-8 · Zbl 1383.92015 · doi:10.1016/0893-6080(89)90020-8
[2] DOI: 10.1109/72.392252 · doi:10.1109/72.392252
[3] DOI: 10.1109/21.199466 · doi:10.1109/21.199466
[4] DOI: 10.1109/72.159070 · doi:10.1109/72.159070
[5] DOI: 10.1109/TSMC.1971.4308320 · doi:10.1109/TSMC.1971.4308320
[6] Farrow SJ., Marcel Dekker. New York pp 1– (1984)
[7] DOI: 10.1016/0925-4005(95)01721-6 · doi:10.1016/0925-4005(95)01721-6
[8] DOI: 10.1016/0045-7906(95)00033-X · doi:10.1016/0045-7906(95)00033-X
[9] DOI: 10.1016/0165-0114(84)90015-0 · Zbl 0554.93070 · doi:10.1016/0165-0114(84)90015-0
[10] DOI: 10.1016/S0165-0114(98)00174-2 · Zbl 0965.93045 · doi:10.1016/S0165-0114(98)00174-2
[11] Oh S. K., Tran. KIEE. 49 (3) pp 145– (2000)
[12] DOI: 10.1109/FUZZY.1993.327364 · doi:10.1109/FUZZY.1993.327364
[13] DOI: 10.1126/science.267326 · Zbl 1383.92036 · doi:10.1126/science.267326
[14] DOI: 10.1016/0165-0114(80)90059-7 · Zbl 0453.93003 · doi:10.1016/0165-0114(80)90059-7
[15] DOI: 10.1109/TSMC.1987.289361 · Zbl 0643.93018 · doi:10.1109/TSMC.1987.289361
[16] DOI: 10.1016/S0020-0255(97)10083-4 · doi:10.1016/S0020-0255(97)10083-4
[17] DOI: 10.1109/91.388173 · doi:10.1109/91.388173
[18] DOI: 10.1109/5326.669559 · doi:10.1109/5326.669559
[19] DOI: 10.1016/S0020-0255(98)10026-9 · Zbl 0953.68590 · doi:10.1016/S0020-0255(98)10026-9
[20] Oh S. K., 9th IFSA World Congress pp 1690– (2001)
[21] Park B. J., 9th IFSA World Congress pp 1361– (2001)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.