ABSTRACT
Y.L. Han State Nuclear Security Technology Center, Beijing, China
W. Ma National Quality Inspection and Testing Center for Surveying and Mapping Products, Beijing, China
ABSTRACT: The research on the forecast of social security incidents covers many subjects such as social science, mathematics and computer science. This paper proposes dividing forecasting and pre-warning of social security incidents into two parts: macroscopic forecast and microscopic forecast, mainly comprehensively reviews the macroscopic forecast method and systemically generalizes several forecast methods such as: time series analysis; regression analysis and data mining technique, etc. The paper also generalizes the application scope, advantages as well as disadvantages of various methods and indicates the research direction and hot spot of this field by combining the features of various types of social security incidents based on the analysis of the mathematical characteristics of each method.
