Predicting the Spatial-Temporal Distribution of Human Brucellosis in Europe Based on Convolutional Long Short-Term Memory Network

Abstract

Brucellosis is a chronic infectious disease caused by brucellae or other bacteria directly invading human body. Brucellosis presents the aggregation characteristics and periodic law of infectious diseases in temporal and spatial distribution. Taking major European countries as an example, this study established the temporal and spatial distribution sequence of brucellosis, analyzed the temporal and spatial distribution characteristics of brucellosis, and quantitatively predicted its epidemic law by using different traditional or machine learning models. This paper indicates that the epidemic of brucellosis in major European countries has statistical periodic characteristics, and in the same cycle, brucellosis has the characteristics of piecewise trend. Through the comparison of the prediction results of the three models, it is found that the prediction effect of long short-term memory and convolutional long short-term memory models is better than autoregressive integrated moving average model. The first mock exam using Conv layer and data vectorizations predicted that the convolutional long short-term memory model outperformed the traditional long short-term memory model. Compared with the monthly scale, the prediction of the trend stage of brucellosis can achieve better results under the single model prediction. These findings will help understand the development trend and liquidity characteristics of brucellosis, provide corresponding scientific basis and decision support for potential risk assessment and brucellosis epidemic prevention and control, and reduce the loss of life and property.