부산대학교 도서관

In this paper, the multi-agency heterogeneous combination forecasting model is used to forecast and analyze the unemployment rate of urban residents in China, based on the data of the National Bureau of Statistics from 2010 to 2019 and the monthly unemployment rate data before May 2023 as the training set and the test set respectively. By selecting important influencing factors and considering the seasonal, periodic and slightly unbalanced characteristics of the data, the missing data are filled by polynomial interpolation method. According to the specific characteristics of references and data, we decided to choose ARIMA model and support vector regression model as a combination, and designed a new weighted exponential average method based on various weighted average methods to carry out weighted synthesis on the prediction results of the above single model. The fitting effect of the model on the test set has passed the test, which has high reference value. Therefore, according to the results, the combined forecasting model proposed in this paper has remarkable accuracy and reliability in the field of urban residents' unemployment rate forecasting in China.