Forest fire driving factor identification and prediction method, system, equipment and medium

The invention relates to a forest fire driving factor identification and prediction method, system and device, and a medium, and the method comprises the steps: creating a random point according to forest fire data, determining the time of the random point when the random point is created, and enabling a fire point and the random point to form complete point distribution data; combining the collected driving factor data with the complete point distribution data to obtain sample data, and randomly dividing the sample data to obtain a plurality of intermediate models; performing learning training on the pre-constructed model based on the intermediate model; identifying important driving factors of each model; and carrying out significance test on the driving factors of each model, and selecting one model from the final models after all the models select the reserved important driving factors to carry out forest fire prediction. According to the method, multiple driving factors influencing forest fire occurrence