Rolling Prediction of Emergency Supplies Based on Support Vector Machine

This study investigates a rolling prediction method of emergency supplies based on post-disaster multisource time-varying information to ensure the prediction accuracy. The proposed method uses historical case, real-time disaster, and time-sharing simulation data as the source data. The method implements attribute reduction of original data samples based on rough set theory and predicts cumulative death tolls in each rolling period by using the rolling time-domain as the basic framework and combining support vector machine (SVM). Then the proposed method estimates material demands in the corresponding period by using the material demand model according to prediction results in a single period. Finally, the proposed method is verified by an experiment with a general mean prediction error of 10.96%. Moreover, the general mean prediction error of methods in the references is 14.13%. Overall, the method has high accuracy and strong timeliness. The results demonstrate that the proposed method has certain scientificity and can provide reliable decision references for the decision-making department after the occurrence of natural disasters.