A fast learnt fuzzy neural network for huge scale discrete data function approximation and prediction

  • Authors:

  • Omid Khayat;Javad Razjouyan;Fereidoun Nowshiravan Rahatabad;Hadi Chahkandi Nejad

  • Affiliations:

  • Young Researchers Club, South Tehran Branch, Islamic Azad University, Tehran, Iran;Garmsar Branch, Islamic Azad University, Garmsar, Iran;Department of Biomedical Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran;Birjand Branch, Islamic Azad University, Birjand, Iran

  • Venue:
  • Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

In real world dataset, there are often large amount of discrete data that the concern is the interpolation and/or extrapolation by an approximation tool. Therefore, a training process will be actually used for definition and construction of the approximator parameters. Huge amount of data may lead to high computation time and a time consuming training process. To this concern a fast learnt fuzzy neural network as a robust function approximator and predictor is proposed in this paper. The learning procedure and the structure of the network is described in detail. Simplicity and fast learning process are the main features of the proposing Self-Organizing Fuzzy Neural Network SOFNN, which automates structure and parameter identification simultaneously based on input-target samples. First, without need of clustering, the initial structure of the network with the specified number of rules is established, and then a training process based on the error of other training samples is applied to obtain a more precision model. After the network structure is identified, an optimization process based on the known error criteria is performed to optimize the obtained parameter set of the premise parts and the consequent parts. At the end, comprehensive comparisons are made with other approaches to demonstrate that the proposed algorithm is superior in term of compact structure, convergence speed, memory usage and learning efficiency.