A stacking-based ensemble learning method for earthquake casualty prediction

Abstract The estimation of the loss and prediction of the casualties in earthquake-stricken areas are vital for making rapid and accurate decisions during rescue efforts. The number of casualties is determined by various factors, necessitating a comprehensive system for earthquake-casualty prediction. To obtain accurate prediction results, an effective prediction method based on stacking ensemble learning and improved swarm intelligence algorithm is proposed in this study, which comprises three parts: (1) applying multiple base learners for training, (2) using a stacking strategy to integrate the results generated by multiple base learners to obtain the final prediction results, and (3) developing an improved swarm intelligence algorithm to optimize the key parameters in the prediction model. To verify the effectiveness of the model, we collected data pertaining to earthquake destruction from 1966 to 2017 in China. Experiments were conducted to compare the proposed method with popular machine learning methods. It was found that the stacking ensemble learning method can effectively integrate the prediction results of the base learner to improve the performance of the model, and the improved swarm intelligence algorithm can further improve the prediction accuracy. Moreover, the importance of each feature was evaluated, which has important implications for future work such as casualty prevention and rescue during earthquakes.