Title :
Application and Realization of Combined Neural Network in Architecture Design of Information Systems
Author :
Xiaomin, Xu ; Ju, Hu
Author_Institution :
Sch. of Inf. Manage., Beijing Inf. Sci. & Technol. Univ., Beijing, China
Abstract :
Being based strongly on experience, while weekly on theories characterizes the architecture design of information systems, In this paper, by adopting SOM and BP neural network techniques, a multi-layer combined architecture of neural networks is put forward and realized, in the pattern-based methods of comprehensive design of information system; consequently a scientific and effective approach is searched to design the information system, as a contributive attempt to combine the artificial intelligence and information system development.
Keywords :
backpropagation; information systems; multilayer perceptrons; self-organising feature maps; software architecture; BP neural network technique; SOM; artificial intelligence; combined neural network; information system architecture design; multilayer neural network; pattern-based method; Artificial intelligence; Artificial neural networks; Computer architecture; Design methodology; Information management; Information science; Information systems; Management information systems; Multi-layer neural network; Neural networks; BP; SOM; architecture design; information systems;
Conference_Titel :
Information Engineering, 2009. ICIE '09. WASE International Conference on
Conference_Location :
Taiyuan, Shanxi
Print_ISBN :
978-0-7695-3679-8
DOI :
10.1109/ICIE.2009.236
