Graduates' Employment Forecasting

In The continuously increasing enrolment of higher education has created some employment problems. It is very meaningful that we scientifically forecast and analysis the graduates employment trends. Based on the basic principle of Grey Model with First Order Differential Equation and one Variable (GM(1,1)) and Power Function x   Transformation, in this paper, a new Variable Weights Buffer First-entry none linear Grey Model with First Order Differential Equation and one Variable (VWBFGM(1,1r)) model was established by minimum total residual sum of square. Furthermore, a VWBFGM(1,1r) (Variable Weights Buffer First-entry Grey Model with First Order Differential Equation and one Variable)BPNN (Back-Propagation) neural network model is established. The single variable change rate of the proposed VWBFGM(1,1r) BPNN model is less than or greater than the index component, can be used for fitting a single sequence of non-negative data and forecasting widely, and also can interfuse the advantages in the uncertainty domain in radial basis function neural network. KeywordsVWBFGM (1,1r) –BPNN Model, University Graduates' Employment Forecasting, Power Function x   Transformation