Optimization Of Drying Kinetics, Physical Properties, And Predictive Modeling of Pumpkin (Cucurbita Maxima) Powder: A Comparative Analysis of Mathematical and Ann Approaches

This study investigates the drying kinetics, physical properties, and predictive modelling of pumpkin (Cucurbitamaxima) powder using both mathematical and artificial neural network (ANN) approaches. Pumpkin, a richsource of beta-carotene and antioxidants, holds significant potential as a functional food ingredient. The dryingprocess was performed at three temperatures (65°C,70°C,75°C) using a tray dryer to examine its effects onmoisture content, drying rates, and product quality. The results revealed that higher drying temperaturesaccelerated moisture reduction but also caused nutrient degradation and texture changes. Among the testedtemperatures, 70°C provided the best balance between efficiency and quality, while 75°C achieved the fastestdrying rate. Mathematical modelling using five drying models, including the Logarithmic and Diffusion models,demonstrated excellent correlation coefficients (R²), indicating their efficacy in predicting drying kinetics