Insurance, Development, and Conservation: Models Based on Time Series Forecasting and Recursive Forecasting Methods Summary
DOI:
https://doi.org/10.71222/ywtvvq60Keywords:
time series forecasting, recursive forecasting, land environment assessment, conservation of historical architecture, insurance prices, insurance client retention rateAbstract
The natural environment is undergoing significant changes due to human activities, with extreme weather profoundly impacting various aspects of life. In response to increasingly frequent natural disasters, insurance companies and land developers require effective mathematical models to balance risk and reward. Meanwhile, communities need a quantifiable scientific approach to preserve properties of cultural, historical, or economic value. For the first question, we use the insurance client retention rate and the annual probability of extreme weather in Japan and the U.S. to build a time series forecasting model. By comparing future insurance retention rates and extreme weather probabilities with current homeowner insurance amounts, we develop a recursive forecasting model to predict future insurance price trends. For the second question, as land cost is the largest expense in real estate development, we analyze an undeveloped site in Hong Kong’s Falling North area using data from the Environmental Protection Agency. Since the latest environmental assessment report is outdated, we recalculated pollutant concentrations and estimated an 8-year government treatment period. We then collected data on heavy rain, typhoons, land prices, and Hong Kong’s insurance retention rate from relevant sources. Using a time series forecasting model, we project these factors over the next eight years and apply a recursive forecasting model to estimate future housing insurance prices. For the third question, we obtained historical data on sandstorms, extreme temperatures, tourist numbers, and revenue from the Forbidden City in Beijing. We forecast sandstorms, extreme weather, and tourist numbers over the next five years, then built a recursive forecasting model incorporating 2023 revenue data to predict future earnings. Finally, we proposed preservation strategies for historic buildings within the Forbidden City.
References
1. E. I. Ionuț et al., “Cloud Optimized Raster Encoding (CORE): A Web-Native Steamtable Format for Large Environmental Time Series,” Geomatics, vol. 1, no. 3, 2021, doi: 10.3390/GEOMATICS1030021.
2. I. Deeva, O. N. Nikitin, and V. A. Kalyuzhny, “Pattern Recognition in Non-Stationary Environmental Time Series Using Sparse Regression,” Procedia Compute. Sci., vol. 156, 2019, doi: 10.1016/j.procs.2019.08.212.
3. D. Kbaier Ben Ismail, P. Lazure and I. Puillat, "Advanced Spectral Analysis and Cross Correlation Based on the Empirical Mode Decomposition: Application to the Environmental Time Series," in IEEE Geoscience and Remote Sensing Letters, vol. 12, no. 9, pp. 1968-1972, Sept. 2015, doi: 10.1109/LGRS.2015.2441374.
4. S. Hashimoto and K. Takeuchi, “Simplification and Accurate Implementation of State Evolution Recursion for Conjugate Gradient,” IEICE Trans. Fund am. Electron. Commun. Compute. Sci., vol. 106, no. 6, pp. 952–956, 2023, doi: 10.1587/TRANSFUN.2022EAL2088.
5. T. L. Miller and H. Sande, “Is word-level recursion actually recursion?” Languages, vol. 6, no. 2, p. 100, 2021, doi: 10.3390/LANGUAGES6020100.
6. R. R. Narayanan, N. Durga, and S. Nagalakshmi, “Impact of Artificial Intelligence (AI) on Drug Discovery and Product De-velopment,” Indian J. Pharm. Educ. Res., vol. 56, pp. S387–S397, 2022, doi: 10.5530/ijper.56.3s.146.
7. S. Yang, J. Hu, P. Gao, R. Dong and Z. Wu, "Recursive Filtering for Time-Varying Nonlinear Delayed Systems with Stochastic Parameter Matrices and Censored Measurements," 2023 42nd Chinese Control Conference (CCC), Tianjin, China, 2023, pp. 1137-1141, doi: 10.23919/CCC58697.2023.10240531.
8. S. SAKUMA, R. SATO, M. NAKAMURA and K. YAMAGOE, "Method for Extracting Suspected Faulty Equipment Through Recursive Use of GNN Model," 2022 23rd Asia-Pacific Network Operations and Management Symposium (APNOMS), Takamatsu, Japan, 2022, pp. 1-6, doi: 10.23919/APNOMS56106.2022.9919900.
