Deep neural networks with application in predicting the spread of avian influenza through disease-informed neural networks

Deep learning has emerged in many fields in recent times where neural networks are used to learn and understand data. This study combined deep learning frameworks with epidemiological models and was aimed specifically at the creation and testing of disease-informed neural networks (DINNs) with a view of modeling the infection dynamics of epidemics. Our research thus trained the DINN on synthetic data derived from a susceptible infected-susceptible infected removed (SI-SIR) model designed for avian influenza and showed the model's accuracy in predicting extinction and persistence conditions. In the method, a twelve hidden layer model was constructed with sixty-four neurons per layer and the rectified linear unit activation function was used. The network was trained to predict the time evolution of five state variables for birds and humans over 50,000 epochs. The overall loss minimized to 0.000006, characterized by a combination of data and physics losses, enabling the DINN to follow the differential equations describing the disease progression.