Using Digital Surveillance Tools for Near Real-Time Mapping of the Risk of International Infectious Disease Spread: Ebola as a Case Study

Sangeeta Bhatia; Britta Lassmann; Emily Cohn; Malwina Carrion; Moritz U.G. Kraemer; Mark Herringer; John Brownstein; Larry Madoff; Anne Cori; Pierre Nouvellet

• If an inconsistent point was at the end of the the cumulative case series, applying the above rules led to the removal of a large number of points. Hence, to remove outliers at the end, we used Chebyshev inequality with sample mean [36] . Given a set of observations X 1 , X 2 . . . X n , the formulation of Chebyshev inequality by Saw et al. gives the probability that the observation X n+1 is within given sample standard deviations of the sample mean. We defined X n+1 to be an outlier if the probability of observing this point given observations X 1 , X 2 . . . X n is less than 0.5. Fixing this probability allowed us to determine k such that P r(µ−kσ ≤ X {n + 1} ≤ µ + kσ) ≥ 0.5, where µ and σ are the sample mean and sample standard deviation respectively. We deleted an observation X n+1 as an outlier if it did lie in this interval.