Research on Disaster Insurance Based on Long Short Term Memory Neural Network (LSTM) and ARAM

With climate change and extreme weather events increasing every year, in order to address the dilemmas faced by industries such as insurance and real estate, in this paper, a catastrophe model and a historic building assessment model are developed. Firstly, the destructive power of disaster-causing factors was predicted using the LSTM model, and then the catastrophe model was constructed by introducing the Area Risk Assessment Model (ARAM) and the P Underwriting Risk Model (PURM). Then, the Historical Building Evaluation Model (HBEM) was established, combining the Delphi method with the relatively subjective cultural and artistic values to obtain more objective quantitative results through the AHP method. Finally, the Longmen Grottoes in China were selected as a historical landmark to verify the model proposed in this paper.