Under the background of globalization, natural disasters pose a challenge to the insurance industry, and an effective property underwriting strategy is urgently needed. Considering these issues, this study collected data on nine indicators related to disaster and property insurance in Turkey, France, India, Australia, and Gabon from websites such as EARTHDATE and predicted them backward by LSTM for five years. Using these data, this paper then built a Disaster Assessment model to determine whether insurance companies should choose to underwrite policies. DA model reflected the overall disaster degree from the three dimensions: population, construction, and economy. Firstly, the entropy weight method is used to calculate the weights of 9 secondary indicators. Then use the K-means clustering algorithm to divide the three-dimensional clustering into 8 boxes. Insurance companies will not underwrite policies in the hard-hit areas. Eventually, apply the DA model to Yonghe Street and Cote de Nuits.

Yu Zheng. Multidimensional Analysis of Insurance Companies' Underwriting Choices: Case Studies of Yonghe Street and Cote de Nuits. Academic Journal of Business & Management (2024) Vol. 6, Issue 7: 167-174. https://doi.org/10.25236/AJBM.2024.060723.

[1] Zhang Peng, Zhang Yunxia, Sun Zhou, et al. Comprehensive disaster index: a quantitative evaluation method of natural disaster loss[J]. Catastrophology, 2015, 30(04):74-78.

[2] Liu Yihang, Shen Hangxian. Spatiotemporal prediction of Traffic Accident Severity based on Ensemble Learning Model [J]. Science and Technology Innovation and Application, 2024, 14(08):28-35.

[3] Zhang Xing, Zhang Chungui, Wu Juxin. Research on weight determination method in comprehensive assessment of agro-meteorological disasters[J]. Chinese Agricultural Science Bulletin, 2008(11):448-452.

[4] Xia Guangwei, He Qingquan, Li Yushan. Research on Financial customer risk based on K-means data analysis [J]. Journal of Insurance Vocational College, 2022, 36(05):52-57.

[5] Wang Chunjing, Liu Dan, Zhang Wenkang, et al. Based on the grade of flotation LSTM neural network prediction model [J]. Journal of chemical industry automation and instrumentation, 2024 ploidy (02):325-332.

[6] Xiao-Ting W , Ai-Hong L , Pei-Long S .Research on Regional Financial Risk Measurement and Evaluation Based on Macro Balance Sheet: Taking Shanxi Province as an Example[J].On Economic Problems, 2019.

[7] Huang Zhimin, Liang Chengdong. Grade Evaluation data Analysis based on K-means clustering algorithm [J]. Electronic Quality, 2023(12):40-44.