Prediction and analysis of skin damage of traditional building in southern Fujian based on machine learning

Traditional folk houses in southern Fujian reflect the construction technology and cultural attributes of residential buildings in southern Fujian, and are an important carrier to study the production and life style in ancient southern Fujian. Due to the adverse external environment such as rainfall and strong wind, the damage of the building skin has become more and more obvious, which has seriously threatened the overall structural stability of the traditional folk houses in southern Fujian. Based on the field investigation data and combined with the surrounding physical environment factors of traditional dwellings in southern Fujian, this paper uses FCM clustering and Gauss process multi-classification methods to make a probability prediction of the development trend of building skin damage diseases, which provides a certain theoretical basis for the protection of traditional dwellings in southern Fujian. The research results show that the data simulation method proposed in this paper can effectively increase the number of samples, and the characteristics of the simulated data are basically consistent with the field investigation data; The median value of the statistical results of 50 randomized trials is about 0.93, indicating that the model established based on the simulated data has good generalization performance, and is suitable for predicting the development trend of skin damage of traditional residential buildings in southern Fujian.