Forecast modeling for structural equation model

Based on the historical data, a forecast modeling method for structural equation model was discussed, where the future relationship between the system factors was described without future sample. By applying spectra of matrix, the covariance matrix was decomposed of eigenvectors and eigenvalues. Typical linear regression method was adopted to predict eigenvalues, and predictive method of orthonomal matrix based on rotations of principal axes was adopted to predict eigenvector matrix, so it structured a forecast method of covariance matrix. The maximum likelihood method was applied to estimate the parameters of future structural equation model. The experimental simulation illustrated main computational procedures of the predictive model.The results show a high precise of the predictive values. The agreement of the final computation results with the experimental data indicates this method could be used to analyze and forecast structural equation model.