Author: Sarah F. McGough; Michael A. Johansson; Marc Lipsitch; Nicolas A. Menzies
Title: Nowcasting by Bayesian Smoothing: A flexible, generalizable model for real-time epidemic tracking Document date: 2019_6_7
ID: 6kq0ptlg_27
Snippet: We introduce a new approach for Bayesian nowcasting and demonstrate its application in two disease contexts with different reporting systems, outperforming an existing method in terms of point estimate (reduced RMSE) and probabilistic (higher logarithmic score) predictive performance. In particular, NobBS performs well even when the delays in case reporting change over time. We further demonstrate an important trade-off related to moving window s.....
Document: We introduce a new approach for Bayesian nowcasting and demonstrate its application in two disease contexts with different reporting systems, outperforming an existing method in terms of point estimate (reduced RMSE) and probabilistic (higher logarithmic score) predictive performance. In particular, NobBS performs well even when the delays in case reporting change over time. We further demonstrate an important trade-off related to moving window sizes for delay distribution estimates; short windows improve the real-time characterization of the delay distribution but are susceptible to over-estimating that variability, potentially decreasing nowcast accuracy. Lacking any disease-specific parameterization, and relying only on historical trends of case reporting as input, this approach can be immediately adapted in a variety of disease settings.
Search related documents:
Co phrase search for related documents- case report and delay distribution: 1, 2, 3
- case report and delay distribution estimate: 1
- case report and disease setting: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21
- delay distribution and disease contexts application: 1
- delay distribution and disease specific parameterization: 1
Co phrase search for related documents, hyperlinks ordered by date