Extreme Weather Prediction and Prevention——Research Based on Machine Learning Method

In this paper, machine learning method is used to predict and prevent extreme weather. Firstly, this paper introduces entropy weight -TOPSIS evaluation method and K-means clustering method, and combines LSTM model to evaluate and predict extreme weather conditions. By collecting a large number of eco-environmental data, this paper analyzes the trend and causes of extreme weather changes in China, and discusses it from the perspective of precipitation and land cover. The research results identify six areas with the most severe extreme weather in the future, and provide specific forecast scores. These findings are of great significance for preparing in advance and reducing the negative effects of extreme weather, which can help to take preventive measures, raise public awareness, strengthen infrastructure construction and optimize resource allocation.