Font Size: a A A

Analysis Of Employment In Shandong Province Based On Artificial Neural Network

Posted on:2021-05-03Degree:MasterType:Thesis
Country:ChinaCandidate:X T WeiFull Text:PDF
GTID:2427330602966297Subject:Applied Statistics
Abstract/Summary:PDF Full Text Request
Shandong Province is a populous province in China.With the continuous growth of population,the employment situation is getting more and more severe.Therefore,how to analyze the number of employees and accurately predict the number of employees is particularly important.This paper uses Pearson correlation coefficient and principal component analysis to select the factors that affect the number of employees in Shandong Province,constructs the BP neural network model to predict the number of employees,and compares the prediction results with the residual auto-regressive model.This paper uses BP neural network to predict and analyze the employment in Shandong Province.This paper selects the annual data of relevant indicators from2000 to 2018,and determines the total population,output value of primary industry,output value of secondary industry,output value of tertiary industry,total investment,GDP per capita,number of persons participated in urban basic pension insurance,the consumer price index,disposable income of urban residents,the income of rural residents,and graduates on regular higher education,a total of 11 indicators to study its impact on employment in Shandong Province.Firstly,the article uses Pearson correlation coefficients to analyze the influencing factors,and uses the variable data from 2000 to 2015 as the training set to establish the BP neural network model.The average relative error of the model training results is0.14%.The model was used to forecast the number of employed persons in Shandong Province from 2016 to 2018.By calculating the relative error between the predicted value and the true value,the relative errors were less than 8%,the average relative error was 3%,and the relative error of the predicted value fluctuated slightly.Secondly,the principal component analysis method was used totransform the 11 original variables into four new comprehensive variables as input variables of the BP neural network to reconstruct the model.The average relative error of the model training results was 0.25%.From 2016 to 2018,the relative error of the employment forecast result is within 5%,and the average relative error is 2.2%.Finally,by comparing the results of predicting and analyzing the number of employed people in Shandong Province by using the residual auto-regressive model,it is concluded that the BP neural network method has better training effect.The method of combining principal component analysis and BP neural network has a relatively good prediction effect on the employment in Shandong Province,and the average error is the smallest.Both methods are superior to the residual auto-regressive model,thus verifying the effectiveness of the BP neural network model in predicting employment.
Keywords/Search Tags:Employment forecast, BP neural network, Principal component analysis, Residual auto-regressive model
PDF Full Text Request
Related items