Research on the Optimization of Insurance Decision-Making Model Based on CRITIC-Entropy Method and Improved K-means Algorithm

This paper focuses on the optimization of an insurance decision-making model based on the CRITIC-entropy method and the improved K-means algorithm. This paper first collects big data related to the insurance industry and economic loss data caused by climate from various regions in China from 2013 to 2023, and processes the data through preprocessing and normalization to ensure the validity of the analysis. Subsequently, a regional meteorological disaster risk assessment model is constructed, employing information diffusion theory to assess the risks associated with extreme weather events. Additionally, a corporate decision-making model is proposed, considering the impact of internal and external risks on insurance premiums, and a multi-period dynamic model is introduced to evaluate the long-term operational capability of insurance companies. The paper further uses the CRITIC-entropy method to determine the weights of the evaluation indicators and combines the improved TOPSIS method to construct the SEIC evaluation indicator model, which comprehensively assesses the performance and response capabilities of enterprises during extreme climate events. Finally, the model is expanded into a multi-period dynamic model, and the regional climate change risk is measured using the K-means algorithm. The BP neural network model is used to predict climate changes to verify the accuracy and effectiveness of the model.