Evaluating and predicting the impact of storage conditions and packaging materials on the physical properties of paddy rice using machine learning approaches and artificial neural networks

The overall quality of paddy rice grains is affected by storage conditions and the environment. To explore this concept, the study examines the influences of storage conditions and bag materials on paddy rice grains' physical properties and quality over 6 months. Three types of storage bags, including Jute bags (JB), Polypropylene woven bags (PWB), and Perforated-polypropylene woven bags (P-PWB), were tested under warehouse (WH: 25 ± 2 °C) and refrigerator (R:4 ± 1 °C) conditions for storing grains at moisture levels of 18, 14, and 12 %. The quality assessment included the evaluation of fungal colonization, whiteness degree, crack percentage, hardness, color, broken percentage, and germination rate. More so, Artificial Neural Networks (ANN) and optimization models were utilized to predict and improve the impact of storage time, storage conditions, and bag materials on the physical qualities of paddy rice. The findings indicate that storing paddy rice at 4 °C with 12 % moisture content in jute bags or P-PWB limits fungal growth and preserves rice quality. P-PWB showed the lowest discoloration percentages across all moisture levels, temperatures, and storage durations, demonstrating its superior protection. For shorter storage, P-PWB is ideal for preserving grain hardness and reducing breakage. Additionally, P-PWB at 4 °C and 12 % MC is highly effective in retaining whiteness and overall visual quality during prolonged storage. Pearson correlation analysis revealed a strong positive correlation between crack ratio and broken percentage, emphasizing the need to maintain grain integrity as cracking increases breakage. Principal Component Analysis (PCA) identified crack ratio, discoloration, and fungal contamination as key factors affecting overall grain quality, with germination, whiteness, and hardness playing supporting roles. More so, integrating Artificial Neural Networks (ANNs) into evaluating and predicting the influence of storage environments and bag materials on paddy rice quality offers a promising approach to enhancing food preservation strategies.