Neuro-Fuzzy Systems Modeling Tools for Bacterial Growth

Many techniques have been used in classification of bacterial growth-non-growth database are network based. This paper proposes adaptive neuro-fuzzy system for classifying the bacterial growth/non-growth and modeling the growth history. A brief description of the neuro-fuzzy intelligent systems scheme is proposed. The performance of neuro-fuzzy system is investigated for their quality and accuracy in classification of growth/no-growth state of a pathogenic Escherichia coli R31 in response to temperature and water activity. A comparison with the most common used statistics and data mining classifiers was carried out. The neuro-fuzzy system classifier was found to do better than both linear/nonlinear regression and multilayer neural networks. Results show bright future in implementing it in food science and medical industry.