As a country with high frequency of geological disasters, China still faces many problems in dealing with geological disasters and reducing economic losses. In recent years, with the deepening of the research on geological disaster early warning, as well as exchanges with various institutions and learning experience, we have made some achievements in how to carry out geological disaster early warning, which has become a major driving force for social development. Nowadays, people mostly deal with geological disasters with certain human intervention. This paper relies on the “big data analysis” and mathematical conventional statistical analysis in recent years, and carries out a series of experiments to establish the early warning model, and carry out corresponding analysis on the multiple attributes of the model, so as to explore the usability of the early warning model in the future geological disasters. In this paper, the corresponding exploration of geological disasters in recent years is carried out. Correlation analysis and statistical analysis are carried out in the aspects of year, disaster frequency, loss and early warning, and the prediction model is established. It is found that with the development of time, the occurrence times of geological disasters are uncertain, and people pay more attention to the early warning of geological disasters and face the loss caused by geological disasters It will decrease gradually.