Bayesian Network-based GF-City Metropolitan Zone Modelling: an Approach Based on Location Quotient Resource Impulse Response Principle

Abstract

Based on the evolution phenomenon and law of “location quotient”, this paper has studied the mapping relationship between image recognition of GF-City metropolitan area and the impulse response of local industrial resources. Based on the “location quotient” method, the level of local industry specialization in Foshan city was analysed, with the aid of VAR model. With the help of economic logic reasoning method, the VAR regression equation model improved design pattern analysis method was extracted in terms of the typical local industry semantics; a Bayesian network-based VAR model GF-City metropolitan area local industrial resources was constructed together with the impulse response network model. Based on the image recognition evaluation experiment of GF-City metropolitan area, the network combined Bayesian scoring function and search algorithm to carry out learning reasoning, and thus divided into local industrial grammar sub-network, image vocabulary sub-network and mapping relationship sub-network by modularization. According to the trend of node and conditional probability distribution, the image lexical structure and morphological organization characteristics were analysed, and the impulsive local industry resource impulse response knowledge rules were subsequently mined and characterized. Consequently, an example analysis proved the validity and applicability of the modelling method for the knowledge acquisition of the “location quotient” resource impulse response, so as to further analyse and perfect the rapid development of Foshan City with the drive of “GF-City metropolitan area” in the background of GHM Greater Bay Area.