Study On The Real Estate Tax Base Batch Evaluation Based On Extreme Learning Machine Optimized By Genetic Algorithm Model

With China’s real estate tax legislation into the legislative plan of the NPC,real estate tax will soon be levied.In order to ensure the real estate tax levy work smoothly,the establishment of real estate tax and real estate tax assessment system should be carried out as a whole,building a universally applicable framework for evaluating the tax base of real estate becomes of top priority.Based on this,the paper introduces the current situation of the construction of real estate tax base evaluation system in various countries,compares and analyzes the taxation basis,taxation object,evaluation method,evaluation model,evaluation mechanism,evaluation cycle of each country,and we can get inspiration and apply it to the construction of real estate tax base evaluation system in China.China’s urban real estate is large in quantity,wide in distribution,versatile in use,and the real estate prices are constantly changing,therefore,the accurate evaluation of real estate prices has become a very important and difficult task for real estate tax levy,the key is to build an effective,fair and just real estate tax base batch evaluation system.Therefore,this paper discusses the batch evaluation method in the construction of the real estate tax assessment system and choose a new extreme learning machine network.According to the pattern of the evaluation area division,selection of the comparable transaction instance and establishment of the batch evaluation model,aiming at the "fuzzy" and "grey" characteristics of real estate value,the idea of combining grey fuzzy clustering method with extreme learning machine network is put forward.Grey fuzzy clustering is used to extract common features from many characteristics of real estate,and the evaluation areas with the same characteristics are clustered into one category.Based on the classification of different assessment regions,the genetic extreme learning machine evaluation model is constructed to evaluate the real estate price,and the empirical analysis is conducted with the example of Qingdao ordinary residential buildings.From the empirical results,the grey fuzzy clustering method can be used to classify the different residential areas effectively,mass appraisal method of real estate price based on grey fuzzy clustering--extreme learning machine model is helpful to improve the accuracy of the mass appraisal and reduce the running cost of the mass appraisal.