Applying ARIMA Modeling and Machine Learning Techniques to Analyze the Impact of Extreme Weather on Community Buildings

This paper focuses on the impact of extreme weather on insurance companies and community properties, highlighting the application of machine learning algorithms in analysis and prediction. By constructing an insurance risk underwriting prediction model and a building protection model, the market strategies of insurance companies as well as the value and disposal options of community buildings are explored. First, machine learning algorithms are applied to analyze the impact of property damage, casualties, and the number of extreme weather events on the insurance industry using extreme weather data for the United States and Africa from 1980 to 2021. Time-series forecasts are performed using ARIMA models to anticipate loss data for the next five years, and insurance profit and loss models are constructed for in-depth analysis. Next, the risk level of each U.S. state was assessed through gray correlation analysis and multiple linear regressions were applied to analyze metrics related to risk level, such as urban population and forested area. This process demonstrates the effectiveness and accuracy of machine learning in risk assessment.