Biology Reference
In-Depth Information
100
60
20
0
0
5
10
15
20
25
30
Time
Figure 2.4
Log-normal outbreak with 1000 cases. Outbreak shape:
μ
- 2 σ - 0.5.
2.6 Example:
Mimicking a BioALIRT Dataset
2.6.1 Mimicking background Health Data
To illustrate the product of the data simulator, we mimic a set of authentic
multivariate daily counts taken from the BioALIRT program conducted by
the U.S. Defense Advanced Research Projects Agency (DARPA) (Siegrist and
Pavlin 2004). We use a dataset of six series from a single city, where three of
the series are indicators of respiratory symptoms and the other three of gas-
trointestinal symptoms. The series come from three different data sources:
military clinic visit diagnoses, filled military prescriptions, and civilian phy-
sician office visits, all within a particular U.S. city. Figure 2.5 displays the six
series of daily counts over a period of nearly 2 years.
To mimic this six series dataset, we estimate the different explainable pat-
terns using the mimicker, and then generate a mimicked copy of the original
dataset. Figure 2.6 displays the six mimicked series. This dataset is clearly
very similar to the authentic dataset in its overall appearance in terms of
count levels and patterns. To better see the similarity in structure with respect
to the cyclic behavior (seasonality in Resp and DOW in both Resp and GI),
