Artificial intelligence-based model for evaluating the inhibition of <i>Listeria monocytogenes, Staphylococcus aureus</i>, and <i>Escherichia coli</i> in kefir matrix

The present study aimed to inhibit the activity of some foodborne pathogens by probiotic lactic acid bacteria (LAB) in kefir. The antimicrobial effect of probiotic LAB was evaluated by using Artificial Intelligence (AI)-based models, Artificial Neural Network (ANN), and Adaptive Network-based Fuzzy Inference System (ANFIS). The experiment was performed on fermentation day 0, 1, and 2, and storage day 1, 3, 7, and 10 of kefir. The average inhibition results obtained for Listeria monocytogenes, Staphylococcus aureus, and Escherichia coli at training stage was 2.4 log10 CFU/g, 2.0 log10 CFU/g, and 2.4 log10 CFU/g in Artificial Neural Network, respectively, and 2.33 log10 CFU/g, 2.04 log10 CFU/g, and 2.03 log10 CFU/g in Adaptive Network-based Fuzzy Inference System, respectively. The average result obtained in the case of tested LAB was 4.9 log10 CFU/g, 4.8 log10 CFU/g, and 4.9 log10 CFU/g, respectively, in Artificial Neural Network in each organism; while similar result was observed in Adaptive Network-based Fuzzy Inference System. The results indicate that the activity of all targeted foodborne pathogens was reduced during fermentation and storage days by the potential probiotic LAB present in kefir. Based on the experiment, it was concluded that the activity of foodborne pathogens can be inhibited by probiotic LAB in kefir. In addition, it was suggested that probiotic bacteria in kefir are promising bio-controlling agents that can be used in the food industry.