Epidemiological Prediction using Deep Learning

Cited 0 times inthomson ciCited 0 times inthomson ci
Epidemiological Prediction using Deep Learning
Heo, Jaewoong
Lee, Chang Hyeong
Epidemiology; Deep Learning
Issue Date
Graduate School of UNIST
Accurate and real-time epidemic disease prediction plays a significant role in the health system and is of great importance for policy making, vaccine distribution and disease control. From the SIR model by Mckendrick and Kermack in the early 1900s, researchers have developed a various mathematical model to forecast the spread of disease. With all attempt, however, the epidemic prediction has always been an ongoing scientific issue due to the limitation that the current model lacks flexibility or shows poor performance. Owing to the temporal and spatial aspect of epidemiological data, the problem fits into the category of time-series forecasting. To capture both aspects of the data, this paper proposes a combination of recent Deep Leaning models and applies the model to ILI (influenza like illness) data in the United States. Specifically, the graph convolutional network (GCN) model is used to capture the geographical feature of the U.S. regions and the gated recurrent unit (GRU) model is used to capture the temporal dynamics of ILI. The result was compared with the Deep Learning model proposed by other researchers, demonstrating the proposed model outperforms the previous methods.
Department of Mathematical Sciences
Go to Link
Appears in Collections:
Files in This Item:
Epidemiological Prediction using Deep Learning.pdf Download

find_unist can give you direct access to the published full text of this article. (UNISTARs only)

Show full item record


  • mendeley


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.