Based on the daily data of visits for respiratory diseases in two grade A hospitals as well as meteorological factors and air pollution in Fuxin City from December 1, 2020 to November 31, 2021, PCA and RBF neural netw...Based on the daily data of visits for respiratory diseases in two grade A hospitals as well as meteorological factors and air pollution in Fuxin City from December 1, 2020 to November 31, 2021, PCA and RBF neural network were used to study the effects of meteorological factors and air pollution on respiratory diseases and predict them. The results showed that the number of daily visits was the largest in winter(accounting for 62.5%), followed by spring(15.2%), and it was the smallest in autumn(only 6.9%). The correlation between the number of daily visits and meteorological factors was higher than that of air pollution factors, and the correlation with temperature and ozone was the highest. The response coefficient of daily visits to each factor increased first and then decreased within 9 d, and the peak was 4-5 d behind. RBF and PCA-RBF neural network models were established to predict the number of daily visits, and the accuracy was 86.3% and 95.2%, respectively.展开更多
基金Supported by the Scientific Research Project of Liaoning Meteorological Bureau (ZD202208, ZD202257)Science and Technology Research Project of Fuxin Meteorological Bureau (FX2022-11, FX2022-13)。
文摘Based on the daily data of visits for respiratory diseases in two grade A hospitals as well as meteorological factors and air pollution in Fuxin City from December 1, 2020 to November 31, 2021, PCA and RBF neural network were used to study the effects of meteorological factors and air pollution on respiratory diseases and predict them. The results showed that the number of daily visits was the largest in winter(accounting for 62.5%), followed by spring(15.2%), and it was the smallest in autumn(only 6.9%). The correlation between the number of daily visits and meteorological factors was higher than that of air pollution factors, and the correlation with temperature and ozone was the highest. The response coefficient of daily visits to each factor increased first and then decreased within 9 d, and the peak was 4-5 d behind. RBF and PCA-RBF neural network models were established to predict the number of daily visits, and the accuracy was 86.3% and 95.2%, respectively.