• Soaking time is the most critical factor that affects the Hydration kinetics of the SAMPEA Cowpea than temperature. • Conventional breeding results in morphological and hydration variation in cowpea. • Machine Learning models offer higher efficiency in predicting the hydration kinetics of cowpea with reduced error. • Surrogate ML models provide a more accurate prediction, while the box ML model can understand non-linear relationships present in hydration kinetics. • Water absorption of cowpea is important property which influences processing methods such as soaking time, water retention, and product texture. Cowpeas constitute a vital legume crop, playing a key role in contributing to zero hunger and human well-being by serving as a major source of plant-based protein and micronutrients. However, despite their potential in promoting food security, cowpea utilization, particularly in industrial food processing, remains limited due to an absence of comprehensive data on their hydration kinetics and thermodynamics properties. This study presents a novel investigation of hydration kinetics and thermodynamics, using conventional models with Machine Learning (ML) algorithms, including Support Vector Regression, Deep Neural Networks, XGBoost, and Random Forest to enhance prediction accuracy. The results revealed significant variability in hydration characteristics among the hybrids: SAMPEA-21 had the highest hydration index of 129.08%, while SAMPEA-16 had the highest swelling capacity and swelling index of 0.58 mL/seed and 61.93%, respectively. The best fit was obtained from the Peleg model among the conventional models, with an R 2 of 0.992 at 30°C and a significant decrease up to 0.250 at 90°C. For the ML models, Support Vector Regression provides the best prediction ( R 2 of 0.9997 and RMSE of 0.0113 g/g), indicating superior predictions and higher accuracy. Furthermore, principal component analysis revealed distinct morphological behavior. This study presented novel insights into the hydration kinetics of SAMPEA hybrids, with potential applications for improving process efficiency and product quality in the food industry, particularly for sustainable processes such as soaking and cooking. This research established a scientific foundation for optimizing cowpea utilization in food processing and promoting value addition and sustainable agriculture.
Garba et al. (Sun,) studied this question.