Figure 3 | Scientific Reports

Figure 3

From: Simulation and identification of foodborne outbreaks in a large supermarket consumer purchase dataset

Figure 3

Details of a single frozen chicken outbreak and analysis. (a) Epidemic curve of simulated frozen chicken outbreak. Background case proportion = 0.5, delay parameter = 150 days, attack rate = 0.01, outbreak length = 50 days. N = 37 cases. (b) Correct identification of outbreak vehicle using logistic regression to generate odds ratios. Y-axis on log scale. After at least 5 cases are registered, the outbreak analysis algorithm of Fig. 2 is run. Every time a new case is registered, the purchase history of cases are used as covariates and case-controlled by the purchase history of randomly selected customers in the CPD. ORs are calculated for each item, and once an item reaches a pre-determined OR threshold (here 50), it is returned as the outbreak vehicle. The top 5 hits are shown in this figure.

Back to article page