Key points are not available for this paper at this time.
Machine learning methods are innovative tools for enhancing drying processes and preventing food deterioration. It helps drying techniques in reducing postharvest losses, avoiding spoiling, and maintaining a stable garlic supply chain. This study predicted the performance of the microwave drying process for garlic slices under various airflow conditions (0.3, 0.5, and 1.0 m/s) and microwave powers (100, 200, and 300 W) using an artificial neural network. The findings revealed that the ANN model's accuracy demonstrated remarkable correlations between the physicochemical response parameters of the dried garlic slices and the input parameters. However, the results indicated that at 300 W and 0.3 m/s, the lowest water activity of 0.41 and the highest color change were attained at 23.79. Reducing the microwave power and increasing the airflow velocity extended the dehydration time. The values of vitamin C and allicin content rose with airflow velocity but decreased with increasing microwave power. Additionally, rehydration decreased as airflow velocity increased, whereas garlic flavor levels decreased as microwave power increased. The ANN model was highly effective in predicting the following parameters during microwave drying. Overall, the use of ANNs in microwave drying contributes to an improved garlic drying process while preserving the physicochemical properties of garlic slices. • Machine learning offers innovative and reliable solutions for addressing food spoilage. • Microwave drying is an efficient alternative for drying food. • Vitamin C and allicin levels increased with airflow but decreased with higher MP. • The ANN proved to be a robust tool for predicting and optimizing physicochemical.
El‐Mesery et al. (Thu,) studied this question.