Modular Neural Networks: A Survey

Auda, G.; Kamel, Mohamed S.

doi:10.1142/s0129065799000125

Cited by 129 publications

(65 citation statements)

References 101 publications

(173 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The third type of decomposition is made during learning process. Modular neural networks have advantages such as improved generalization, better usability, interpretability, easier hardware implementation (Auda and Kamel 1998a). On the other hand, there exist some computational problems: task-decomposition, training network modules, and combination of network outputs for the global optimal solution.…”

Section: Modular Neural Networkmentioning

confidence: 99%

“…The neural network approach to classification is one of the most competent in terms of performance due to its adaptive learning, high non-linearity, robustness and plausibility for hardware implementation. However, traditional neural network classification methods suffer from the inefficiency and the inability of the learning algorithm to converge when the classification problems are large in size and of high complexity (Auda and Kamel 1998a). To avoid these drawbacks, a cooperative modular neural network (CMNN) approach was developed and successfully applied to multi-class problems.…”

Section: Crnns For Classificationmentioning

confidence: 99%

See 1 more Smart Citation

Cooperative recurrent modular neural networks for constrained optimization: a survey of models and applications

Kamel

Xia

2008

Cogn Neurodyn

View full text Add to dashboard Cite

Constrained optimization problems arise in a wide variety of scientific and engineering applications. Since several single recurrent neural networks when applied to solve constrained optimization problems for real-time engineering applications have shown some limitations, cooperative recurrent neural network approaches have been developed to overcome drawbacks of these single recurrent neural networks. This paper surveys in details work on cooperative recurrent neural networks for solving constrained optimization problems and their engineering applications, and points out their standing models from viewpoint of both convergence to the optimal solution and model complexity. We provide examples and comparisons to shown advantages of these models in the given applications.

show abstract

Section: Modular Neural Networkmentioning

confidence: 99%

Section: Crnns For Classificationmentioning

confidence: 99%

Cooperative recurrent modular neural networks for constrained optimization: a survey of models and applications

Kamel

Xia

2008

Cogn Neurodyn

View full text Add to dashboard Cite

show abstract

“…Different model of neural network have been used for the comparison of the result. Among these models, multilayer perceptron neural network [22] with hidden layer neuron numbers 4, 6 and 10, modular neural network [23] with hidden layer number 2 and fuzzy logic (canfis) neural network [24] [25,26] algorithm. In the present study, neural network software NeuroSolutions v6.02 [27] has been used for the calculations.…”

Section: Artificial Neural Network (Ann)mentioning

confidence: 99%

106Sg, 107Bh and 108Hs çekirdekleri için Alfa Bozunum Enerjilerinin Belirlenmesi

Akkoyun

Bayram

2015

CFD

View full text Add to dashboard Cite

Özet. Hem teorik hem de deneysel bakış açısıyla, super ağır bölgenin özelliklerini anlamak için, buradaki çekirdeklerin alfa bozunum özelliklerini incelemek oldukça ilgi çekicidir. Nükleer fizikteki pek çok olguda başarılı sonuçlar veren yapay sinir ağları (YSA), super ağır çekirdeklerin deneysel alfa bozunum enerjileri üzerinde kullanılmış ve Sg, Bh ve Hs çekirdeklerinin alfa bozunum enerjileri bu yöntemle tahmin edilmiştir. Farklı modeldeki YSA'lar veriler üzerinde test edilmiştir. Bu çalışmanın sonuçları, YSA metodunun bu iş için uygun olduğunu göstermektedir. Sonuçlara ait ortalama kare hatalar, ½1-1,5 civarındadır. En iyi sonuçlar, bulanık mantık YSA modeli ile elde edilmiştir.Anahtar Kelimler: Alfa bozunum enerjisi, super ağır çekirdekler, yapay sinir ağları Determination of Alpha Decay Energies for 106 Sg, 107 Bh and 108 Hs nucleiAbstract. Alpha decay properties of superheavy nuclei are challenging for understanding the properties of superheavy region from both theoretical and experimental side. Artificial neural network (ANN) which is successful in many phenomenon of nuclear physics has been employed for the experimental alpha decay energy data of superheavy nuclei in order to estimate unknown alpha decay energies for Sg, Bh and Hs nuclei. Different models of the ANN have been tested on the data. The results of the present study indicate that ANN is capable in this task. The mean square erros for the estimations are about 1-1.5%. The best result have been obtained by using fuzzy logic ANN model.

show abstract

“…The parameter can be also formulated as the output of rule based system like modular neural network structure (Auda & Kamel, 1999). Figure 7 shows the structure of Modular Fuzzy Model.…”

Section: Model Structurementioning

confidence: 99%

Hierarchical reinforcement learning using a modular fuzzy model for multi-agent problem

Watanabe¹,

Takahashi²

2007

2007 IEEE International Conference on Systems, Man and Cybernetics

View full text Add to dashboard Cite

Modular Neural Networks: A Survey

Cited by 129 publications

References 101 publications

Cooperative recurrent modular neural networks for constrained optimization: a survey of models and applications

Cooperative recurrent modular neural networks for constrained optimization: a survey of models and applications

106Sg, 107Bh and 108Hs çekirdekleri için Alfa Bozunum Enerjilerinin Belirlenmesi

Hierarchical reinforcement learning using a modular fuzzy model for multi-agent problem

Contact Info

Product

Resources

About