Development of prediction software to describe total mesophilic bacteria in spinach using a machine learning-based...

Area of Science:

• Food Microbiology
• Computational Biology
• Data Science

Background:

• Accurate prediction of microbial growth is crucial for food safety and shelf-life assessment.
• Traditional models for predicting bacterial growth have limitations in accuracy and adaptability.

Purpose of the Study:

• To develop and evaluate machine learning (ML) regression models for predicting total mesophilic bacteria growth in spinach.
• To compare the performance of ML models against established predictive models.

Main Methods:

• Support Vector Regression (SVR), Decision Tree Regression (DTR), and Gaussian Process Regression (GPR) were employed as ML models.
• Performance was assessed using statistical indices, including the coefficient of determination (R²) and root mean square error (RMSE).
• ML models were benchmarked against modified Gompertz, Baranyi, and Huang models.

Main Results:

• ML-based regression models demonstrated superior predictive accuracy.
• Achieved R² values of at least 0.960 and RMSE values of at most 0.154.
• ML models significantly outperformed traditional methods in predicting bacterial growth.

Conclusions:

• Machine learning regression models offer a more accurate and reliable approach for predicting total mesophilic bacteria in spinach.
• The developed ML tool serves as a viable alternative to traditional simulation methods in predictive food microbiology.
• This advancement has significant potential for enhancing food safety and quality control through predictive modeling.