A cross-contamination risk assessment model with improved coefficient optimization for Campylobacter

In the stages of poultry slaughter and processing, the individuals infected with Campylobacter would pollute the environment which, in turn, would pollute the uninfected individuals, leading to cross-contamination and increasing risks. We employed the Kolmogorov–Smirnov test, indirectly estimating the coefficients by searching the large-scale coefficient space with genetic algorithm, trying to make the model distribution fit the actual distribution as much as possible by this algorithm and further by gradient descent. Finally, we successfully obtained a set of optimal coefficients, quantitatively making clearer the nature of cross-contamination, and thereupon proposed corresponding suggestions for improving the slaughter and process operations.