Artificial neural networks (ANNs) and multiple linear regression (MLR) for prediction of moisture content for coated pineapple cubes

The effects were investigated of edible coating and drying temperature (50, 65 and 80 degrees C) on the properties of dehydrated pineapple cubes. A comparative study was performed using mathematical models, multiple linear regression (MLR) and artificial neural networks (ANNs) to predict moisture ratio (MR) and drying rate (DR). Kinetic drying, effective moisture diffusion (D-eff) and activation energy were examined. Midilli et al. model was the best fit to predict MR. Higher air temperature reduced drying time and moisture content, while D-eff increased for uncoated and coated dried pineapples of 1.69 x 10(-9) to 5.57 x 10(-9) m(2)/s and 1.60 x 10(-9) to 5.95 x 10(-9) m(2)/s with the activation energy (E-a) of 37.68 and 41.61 kJ/mol, respectively. Interestingly, the edible coating did not significantly affect D-eff and E-a, but it retained ascorbic acid. Moreover, ANNs model was appropriate for the prediction of the MR and DR of dehydrated pineapple cubes, as this model had the highest R-2 and accuracy with the lowest RMSE and MAE. The ANNs models with topology of 3-14-14-1 for MR and 3-7-7-1 for DR predictions were the optimal to estimate the drying process of uncoated and coated fruits with satisfactory accuracy and benefit for food industry.

Automated summary

The effects of edible coating and drying temperature on the properties of dehydrated pineapple cubes were investigated. Mathematical models and artificial neural networks were used for prediction. The results showed that higher drying temperature reduced drying time and moisture content, and edible coating had little effect on moisture diffusion and activation energy. Artificial neural networks models accurately predicted the moisture ratio and drying rate of dehydrated pineapple cubes, benefiting the food industry.