Abstract:
A good web selection algorithm can provide the most suitable service for users. However, known for its slow convergence rate and proneness of oscillation in its learning ...Show MoreMetadata
Abstract:
A good web selection algorithm can provide the most suitable service for users. However, known for its slow convergence rate and proneness of oscillation in its learning process, the traditional error back propagation neural network algorithm cannot be applied in the service selection scenarios of actual smart distribution grid. In order to meet the requirements of telecommunication technology for smart distribution grid and improve the quality of telecommunication service, this paper proposes an improved error back propagation algorithm, in which the learning factor can be self-adjusted with every iteration. The simulation results show an optimization of the training speed and an oscillation reduction in the learning process with the new algorithm, thus obvious optimizing the web services selection in smart distribution grid.
Date of Conference: 17-19 September 2014
Date Added to IEEE Xplore: 29 December 2014
Electronic ISBN:978-4-88552-288-8