Research on Early Warning of Social Stability Based on Two-Dimensional Particle Swarm Algorithm

The concept of “thinking in peace” is a way to measure and warn about social stability, which has been highly valued by visionary rulers and politicians. In this paper, we first established a system of indicators affecting social stability, and the Pearson correlation coefficients indicate that they have strong correlation. Based on this, a machine learning model was established to obtain a supervised learning model for both, Bayesian optimization of SVM vector machines, supervised learning using self-referencing Gaussian medium SVM vector machines, prediction confidence histograms obtained by cross-validation methods, and prediction data brought into the model, the relative error between the true stability coefficient and the regression stability coefficient is small, with an accuracy of 91.8% and RMSE <0.4, with strong rationality and scientific validity. Finally, taking the Cuban color revolution as an example, the two-dimensional particle swarm algorithm was applied to analyze the changes of relevant index data before and after the outbreak of the color revolution, and it was concluded that the social stability should show a gradual decrease, and then the main reasons for the failure of the color revolution were analyzed.