Factors Influencing Analysis And Prediction Of Inter-provincial Migration Based On Multi-source Data

With the continuous advancement of reform and opening up,especially since the beginning of the 21st century,my country's domestic market economy has been deepened and developed,and the process of urbanization in various provinces and regions has accelerated,which has effectively promoted domestic economic development,cultural development,and cultural development.The process of social development and modernization.In this context,the large-scale inter-provincial migration and flow of population continue to have a profound impact on economic development and social development.Therefore,a combination of quantitative and qualitative methods is used to analyze the needs of my country's inter-provincial population migration.And prediction is particularly important.The predictive analysis of inter-provincial population migration is the result of the interaction of many influencing factors.Therefore,the predictive trend of population migration has complicated changes.However,the traditional gravity model,binary choice model,multiple linear regression model and other predictive models often cannot completely affect the population.The migration prediction makes a good reflection,making the experimental model unable to get the desired results;therefore,this article introduces the new quantitative source of network search data,collects different data sources,and builds 4 prediction models:gravity Models,reconstructed population migration models,optimized population migration gravity models,improved multiplicative component models to predict the population migration trends of other provinces in China,Beijing,Shanghai,and Guangdong;this article mainly studies China's inter-provincial population from four aspects Analysis and prediction of migration influencing factors.First,the introduction of Baidu index search data,regression analysis and correlation analysis of Baidu index data collected by regions,confirmed that the Baidu index search data can be migrated to Beijing,Shanghai,and Guangzhou.The impact factor of the region of the country is used for population migration prediction analysis,which makes the research in this direction have research value;second,the factor analysis method for the impact factor data of the migration to Beijing,Shanghai,and Guangdong is obtained by rotating the load matrix.The component score coefficients of each influencing factor are calculated to obtain the weight value between each influencing factor,which is helpful to analyze the changes of each influencing factor on the migration to the north,Shanghai,and Guangzhou areas and the construction of model parameters;third,on the basis of the gravity model,Define the migration model and optimize the model parameters according to the domestic situation to analyze the migration trend of the migration in Beijing,Shanghai and Guangzhou;fourthly,starting from the migration population data itself,construct the migration rate,the migration rate and other data to establish an improved multiplicative component model.The prediction results of four model experiments show that:the gravity model based on the traditional gravity model theory,the reconstructed population migration model and the optimized population migration model:among the three models:the traditional gravity model predicts Beijing,Shanghai,and Guangdong for migration purposes The overall R~2 score of the land is the lowest,the reconstructed population migration model predicts the middle,and the optimized and reconstructed population migration model has the best integrated R~2score.Empirical evidence shows that the combination of Baidu index search data and economic factors affecting migration can be carried out.Population migration is predicted and the results are very accurate;while the improved multiplicative component model model based on population migration data sources defines the migration volume,attraction coefficient,migration ratio,etc.and the number of populations in each province to predict the migration of Beijing,Shanghai,and Guangdong in 2015.For predicting the number of people,the improved multiplicative component model has a high prediction accuracy,which can well meet the requirements of some experiments.