Knowledge Enhanced Deep Learning: Application to Pandemic Prediction

Deep Learning has been successfully applied to many problem domains, yet its advantages have been slow to emerge for time series forecasting. For example, in the well-known M Competitions, until recently, hybrids of traditional statistical or machine learning (e.g., gradient boosting) techniques were the top performers. With the recent architectural advances in deep learning being applied to time series forecasting, such as encoder-decoders with attention, transformers, representation learning, and graph neural networks, deep learning has begun to show its advantages. Still, in the area of pandemic prediction, there remain challenges for deep learning models: the time series is not long enough for effective training, ignorance of accumulated scientific knowledge, and interpretability of the model. Today, there is a vast amount of knowledge available that deep learning models can tap into, including Knowledge Graphs and Large Language Models fine-tuned with scientific domain knowledge. There is ongoing research examining how to utilize or inject knowledge into deep learning models. The state-of-the-art approaches are reviewed and suggestions for further work are provided. Recommendations for how this can be applied to future pandemics are given.