Research On Mode Selection Of PPP Projects Of Pension Real Estate Based On SVM Classification Algorithm

Because of the great change of China’s population composition and the acceleration of China’s aging and fewer children,there are more and more imbalances of supply and demand in the aging market.But as far as reality is concerned,there is a great gap between China’s current supply of social old-age services and the growing demand for old-age services in China.In recent years,our government encourages the private sector to invest in our public services,actively advocates the development of PPP model,and introduces the PPP financing model.On the one hand,the PPP model can strengthen the construction of public service projects,so as to meet the needs of our country’s elderly people,but also to effectively improve the pension real estate projects in the lack of technology,construction funds,operation and management chaos;On the other hand,with the development of the pension industry,the PPP model will further expand the development space.However,with the further development of pension PPP projects,its future will also face many problems,such as pension real estate PPP project mode of operation of choice.First of all,the paper analyzes the related concepts,attributes,characteristics and theories of PPP model,and on this basis elaborates the feasibility of PPP model in pension real estate projects.After that,because there are subjective factors in the selection of the influencing factors,the grey relational analysis method is selected to prove the rationality of the selection of the influencing factors.Then,this paper compares the existing PPP mode selection methods,finally chooses the machine learning method as the research method of this paper,and compares several commonly used algorithm models.Secondly,based on the analysis of the current situation of PPP projects of pension real estate in the PPP project library of the Ministry of Finance,and the characteristics and attributes of pension real estate itself,combined with the previous research results of scholars,this paper will select eight factors related to pension real estate,including totalinvestment,franchise period,industry,return mechanism,VFM indicators,project type,regional gross product(GDP),elderly population(over 65 years old),and then collect and sort out the data indicators we need according to the relevant websites.Finally,on the basis of the theory of Chapter 2,this paper chooses the SVM classification model and the kernel function needed by the model,optimizes and adjusts the parameters of the model,constructs the SVM classification prediction and selection model,processes the data well and then inputs it into the SVM classification model for training.To sum up,this paper constructs the PPP mode selection model of old-age real estate by analyzing the current situation of PPP mode selection,which lays a solid foundation for the PPP project selection.At the same time,it also has certain value in promoting the smooth operation of pension real estate PPP projects.