Construction of BN Using Similarity Networks with Application to Development of Intelligent MADM DSS

Chen Linyi

Abstract

With the rapid development of intelligent multiple attribute decision support system, it can assist decision-makers to make decisions scientifically and efficiently by organically combining qualitative analysis and quantitative analysis, and integrating multiple multi-attribute decision making methods, through appropriate programming and friendly user interface. However, most of the existing Bayesian networks in the intelligent multi-attribute decision support system are constructed manually based on the knowledge and experience of experts in the decision-making field. The construction process is very tedious and is not suitable for solving large-scale problems, which will cause huge obstacles and problems in the expansion and development of the system. In order to solve this problem effectively and be better applied to the development of intelligent multi-attribute decision support system to assist decision making, first of all, based on the research of multi-attribute decision making method and Bayesian network, the Bayesian network of knowledge guidance subsystem of intelligent multi-attribute decision support system is constructed and improved based on the traditional Bayesian network construction method. Secondly, this paper proposes a Bayesian network construction method based on probabilistic similarity network model, which greatly improves the efficiency of model construction and the dynamics and extensibility of Bayesian network, and is more effective for solving large-scale problems. Then, the Bayesian network models constructed by two different methods are compared and analyzed from multiple angles. Finally, this paper uses Bayesian network based on probabilistic similarity network model, solved “New Energy Automobile Brand Choose” multiple attribute decision making problem, gives the result of MADM method selection, proves the validity of the Bayesian network, and the core of knowledge guidance subsystem of intelligent multiple attribute decision support system is realized.