TY - GEN
T1 - Machine learning for bacterial growth prediction and examination of bacterial impact on growth
AU - Ince, Volkan
AU - Bader-El-Den, Mohamed
AU - Arabikhan, Farzad
AU - Sari, Omer Faruk
N1 - Publisher Copyright:
© 2024 IEEE.
PY - 2024/10/9
Y1 - 2024/10/9
N2 - In recent years, the growing concern about bacterial growth has become a significant issue in food safety and quality assurance. In this context, the increasing use of machine learning plays a crucial role in understanding and managing these microbial processes in the food industry. This study conducts a comprehensive analysis of bacterial growth in pork products by integrating machine learning methodologies with the high-throughput sequencing technology known as 16S rRNA metabarcoding, which is commonly used to characterize microbial communities, by estimating the relative abundance of microbes. The research investigates the detection and prediction of the growth of microbial populations on food, crucial for safe-guarding food safety and quality in the industry. Utilizing various machine learning algorithms, the study explores the relationship between bacterial quantities and tryptose-sulfite-cycloserine agar (TSC), which is used for the enumeration of Clostridium organisms in a variety of foods and food ingredients. Moreover, the integration of explainable AI (i.e., SHAP) analysis highlighted specific influential bacterial genera, particularly emphasizing the role of Clostridium in the context of TSC values. Additionally, in the study, feature reduction was performed using SHAP analysis. The initial set of features, consisting of 200 genera, was reduced to 8 genera, and the impact of the model was examined. In the comparison between these two different feature counts, the Multi-Layer Perceptron model achieved the highest the area under the curve value (0.85). The SHAP outcomes indicated that the MLP model heavily depended on the Carnobacterium genus, with the Clostridium genus emerging as the second most crucial genus in the model. This research emphasizes the significance of advanced methodologies in understanding microbial communities, contributing to the development of robust strategies for food safety and quality assurance in the food industry.
AB - In recent years, the growing concern about bacterial growth has become a significant issue in food safety and quality assurance. In this context, the increasing use of machine learning plays a crucial role in understanding and managing these microbial processes in the food industry. This study conducts a comprehensive analysis of bacterial growth in pork products by integrating machine learning methodologies with the high-throughput sequencing technology known as 16S rRNA metabarcoding, which is commonly used to characterize microbial communities, by estimating the relative abundance of microbes. The research investigates the detection and prediction of the growth of microbial populations on food, crucial for safe-guarding food safety and quality in the industry. Utilizing various machine learning algorithms, the study explores the relationship between bacterial quantities and tryptose-sulfite-cycloserine agar (TSC), which is used for the enumeration of Clostridium organisms in a variety of foods and food ingredients. Moreover, the integration of explainable AI (i.e., SHAP) analysis highlighted specific influential bacterial genera, particularly emphasizing the role of Clostridium in the context of TSC values. Additionally, in the study, feature reduction was performed using SHAP analysis. The initial set of features, consisting of 200 genera, was reduced to 8 genera, and the impact of the model was examined. In the comparison between these two different feature counts, the Multi-Layer Perceptron model achieved the highest the area under the curve value (0.85). The SHAP outcomes indicated that the MLP model heavily depended on the Carnobacterium genus, with the Clostridium genus emerging as the second most crucial genus in the model. This research emphasizes the significance of advanced methodologies in understanding microbial communities, contributing to the development of robust strategies for food safety and quality assurance in the food industry.
KW - Explainable AI
KW - Food safety
KW - Food spoilage
KW - Supervised machine learning
UR - http://www.scopus.com/inward/record.url?scp=85208448181&partnerID=8YFLogxK
U2 - 10.1109/IS61756.2024.10705237
DO - 10.1109/IS61756.2024.10705237
M3 - Conference contribution
AN - SCOPUS:85208448181
SN - 9798350350999
T3 - IEEE International Conference on Intelligent Systems
SP - 1
EP - 6
BT - 2024 IEEE 12th International Conference on Intelligent Systems, IS 2024 - Proceedings
A2 - Sgurev, Vassil
A2 - Jotsov, Vladimir
A2 - Piuri, Vincenzo
A2 - Doukovska, Luybka
A2 - Yoshinov, Radoslav
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 12th IEEE International Conference on Intelligent Systems, IS 2024
Y2 - 29 August 2024 through 31 August 2024
ER -