Modeling of Employment Structural Transformations

Abstract

The article presents the modelling of structural transformations in the employment of the population by the algorithm: 1) forming a database and establishing the type of mathematical model; 2) identification of variables; 3) exploring the forms of correlation between factor features and the resulting variable; 4) parameterization of data; 5) drawing up the regression equation; 6) economic interpretation of the results, in particular, the calculation of elasticity coefficients; 8) validation of the model for adequacy (calculation of errors of approximation, correlation coefficients and determination); 9) validation of the model for statistical significance (Fisher criterion and Student's t-test); 10) forecasting: calculation of confidence intervals. It is formed the statistical research base of the level dependence of the Ukrainian population employment (selected regions) on a number of factors (share of full-time workers with higher education in % to accounting number; growth rate of labor productivity; growth rate of average wages; capital investment index; coverage ratio). The research was made due to data of the State Statistics Service of Ukraine in the period of 2000-2016 in the Dnipropetrovsk, Kharkiv, Kyiv and Vinnytsia regions. The dynamics of growth of selected factors for each region for the period 2000-2016 is considered. It is established and determined the economic and mathematical models of the level dependence of the Ukrainian population employment (region) on the influence of selected factors for Dnipropetrovsk, Kharkiv, Kyiv and Vinnitsa regions. The comparative analysis of the obtained results is carried out. It is established that the determinants that influence the dynamics of employment growth for the industrial regions of Dnipropetrovsk and Kharkiv regions are the factors reflecting the share of full-time employees with higher education, growth rate of labor productivity and the coefficient of export coverage. It is made a comparative analysis of the built regional models (Dnipropetrovsk, Kharkiv, Kyiv and Vinnytsia regions) with a mathematical model of multiple regression for the analysis of structural changes in the employment of the population of Ukraine.

Research on personalised recommendation algorithm for college students’ employment

“…Finally, the similarity measurement method is used in the algorithm and the most similar items of the user are found in the clustering as the recommendation candidate set. Ilich et al [15] optimised the algorithm based on dimensionality reduction and clustering technology. The algorithm used K-means algorithm and singular value decomposition (SVD) to complete the similarity for different users, which is beneficial for clustering and dimensionality reduction to process the target.…”