Title :
Recurrent Sigmoid-Wavelet Neurons for Forecasting of Dynamic Systems
Author :
Azeem, Mohammad Fazle ; Banakar, Ahmad
Author_Institution :
Aligarh Muslim Univ., Aligarh
Abstract :
In this paper, recurrent neuron models used in feed-forward network are proposed. Each neuron in this model is composed of the sigmoidal activation function (SAF) and wavelet activation function (WAF). The output of the proposed neuron is the product of output from SAF and WAF. In recurrent neuron models delayed output of the sigmoidal and the wavelet activation function is feedback to each other. Performance of the recurrent models is evaluated on two different kind of benchmark problem of dynamical systems and compared with earlier proposed models.
Keywords :
feedforward neural nets; recurrent neural nets; transfer functions; wavelet transforms; SAF; WAF; dynamic system forecasting; feed-forward network; recurrent sigmoid-wavelet neuron; sigmoidal activation function; wavelet activation function; Artificial neural networks; Feedforward systems; IEEE members; Neural networks; Neurofeedback; Neurons; Nonlinear dynamical systems; Output feedback; Power system modeling; Propagation delay;
Conference_Titel :
Information Reuse and Integration, 2007. IRI 2007. IEEE International Conference on
Conference_Location :
Las Vegas, IL
Print_ISBN :
1-4244-1500-4
Electronic_ISBN :
1-4244-1500-4
DOI :
10.1109/IRI.2007.4296679
