Foreign Student Party Construction Model for the Sparse Large Data Equilibrium

In this paper, we propose a foreign student party construction mode (shorted as FSPCM) for sparse large data equilibrium. In FSPCM, we construct sparse large data based on category tag variables, which not only enables the classifier to conduct the foreign student party construction, which makes the effect of the foreign student party construction of each classifier can be greatly enhanced, and the correlation between the marker classifier and the sparse large data can be established by the foreign student party construction, so that the correlation of the sparse large data obtained by the foreign student party construction is more accurate. Through the use of two different loss functions, logistic regression and least squares. The FSPCM-LR (FSPCM with logistic regression) algorithm and the FSPCM-LS (FSPCM with least squares) algorithm are proposed and extended into the regenerated nuclear Hilbert space. Finally, through the alternation solution, we can get FSPCM-LR and FSPCM-LS, compare with 20 reference data sets, and use five different evaluation criteria. Finally, the results show that FSPCM is better than the proposed sparse large data algorithm.