Optimal design of weigh for networks based on rough sets

  • Authors:

  • Baoxiang Liu;Shasha Hao

  • Affiliations:

  • College of Science, Hebei United University, Tangshan, Hebei, China;College of Science, Hebei United University, Tangshan, Hebei, China

  • Venue:
  • ICICA'11 Proceedings of the Second international conference on Information Computing and Applications
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

When the traditional rough neural network is structured, The selection of initial weights are random values between (0,1).This article address this issue, proposed an application of rough set theory attribute importance, replaced with the attribute importance method of initial weights. Finally, with instance validation, compared to the traditional rough neural network,This method is not only to accelerate the network convergence rate, but also enhances the adaptability of BP neural network.