Skip to main content

Table 1 Detection probabilities (multiplied by 100) by forecast method, length of training data, and data set.

From: Accounting for seasonal patterns in syndromic surveillance data for outbreak detection

  Forecast Method
Years of Training Data 1 Eq. (1) 2 EWMA 3 Avg1 4 Avg2 5 HistAvg1 6 HistAvg2 7 HistAvg3 8 HistAvg4
1 year 88,86,100 93,92,88 92,72,86 89,78,84 81,74,84 84,75,100 71,61,92 79,74,100
2 years 69,79,49 94,91,90 92,73,89 90,78,89 74,61,62 77,65,68 70,53,55 73,63,64
3 years 70,99,95 89,93,93 87,92,91 81,88,89 66,90,65 69,94,82 61,89,69 68,90,66
4 years 83,94,99 89,90,91 89,89,89 81,89,71 79,89,97 78,90,97 76,88,95 66,92,94
5 years 92,90,68 91,95,85 88,92,72 84,91,75 76,83,73 77,86,75 73,80,70 76,83,75
6 years 75,91,96 90,95,86 90,88,82 86,72,66 68,84,93 68,90,93 66,84,93 68,89,90
7 years 96,92,66 93,86,91 92,80,92 90,88,87 97,90,72 97,92,74 96,90,70 96,91,73
8 years 82,93,96 95,96,91 94,91,84 92,89,81 87,90,94 88,92,94 88,88,94 89,87,93
Average 83,91,84 92,92,89 91,85,85 87,84,81 79,88,82 79,91,86 77,86,82 77,89,82
  1. The (x,y,z) entry in each cell denotes 100 times the DP (with a false alarm rate of one per year) for the real data, data simulated from the nonhierarchical model, and data simulated from the hierarchical model, respectively. Individual entries are each based on 1000 simulated outbreaks, so their confidence limits are approximately ± 2. In the last row, boldface entries indicate optimistic DPs attributable to using data simulated from the nonhierarchical model. The eight methods are based on applying Page's test to forecast errors based on forecasts from: (1) a fit to Eq. (1), (2) EWMA, (3) moving average with a 1-day gap, (4) moving average with a 3-day gap, (5) historical average using day d, d-1, d+1 in the training data to predict day d in the testing data, (6) historical average using three 4-day windows, (7) historical average using day d, and the (8) historical average using one 4-day window.