Modeling of coconut milk residue incorporated rice‐corn extrudates properties using multiple linear regression and artificial neural network

Abstract

The effect of extrusion screw speed (200, 250, and 300 rpm), barrel temperature (100, 120, and 140 C), and formulation (Coconut milk residue [CMR] 10-20%, corn flour 20-30% and rice flour 60%) on product characteristics like expansion ratio, bulk density, water solubility and water absorption index, compression force, and cutting strength were investigated using multiple linear regression (MLR) and artificial neural network (ANN). The coefficient of determination (R 2 ) of MLR ranged between 0.34 and 0.84, and the sum of squared error (SSE) ranged between 0.0009 and 292.51. Whereas, the R 2 of ANN ranged between 0.41 and 0.94, and SSE ranged between 0.0001 and 214.81. This indicates its superior performance over MLR in the present study. The extrusion condition of 15% CMR, 25% corn flour, and 60% rice flour, at 220 rpm screw speed, and 140 C barrel temperature were determined as optimum conditions for development of coconut milk residue incorporated rice-corn based extrudates with a desirability value of 0.95 using MLR with optimum responses of expansion ratio 3.19, bulk density 0.08 g/cm 3 , water absorption index 5.69 ml/g, compression force 20.80 N, and cutting strength 10.81 N. Practical applicationsCoconut milk residue, which is rich in dietary fiber and polyphenols, is the main underutilized co-product of virgin coconut oil, coconut milk powder, coconut milk yogurt, and flavored coconut milk processing industries. It can be incorporated into the rice-corn mixture to produce a healthy snack food by extrusion. Hence, this study was focused on optimizing the extrusion conditions and flour ratio using multiple linear regression and artificial neural network to obtain a desirable extruded product. The promising results suggest that CMR can be incorporated with rice and corn to produce extrudates with improved nutrition.

Application of artificial neural networks for predicting parameters of commercial vacuum cooling process of baby cos lettuce

“…Moreover, an ANN model with the two hidden‐layer topology 5‐5‐4‐1 increased R 2 and reduced RMSE and MAE in the training and testing stages and provided a more powerful tool than a regression model for predicting seed yield of safflower ( Carthamus tinctorius L.; Abdipour et al, 2019). The modeling of the extrusion process of coconut milk residue incorporated rice‐corn based indicated coefficient of determination ( R 2 ) of ANN ranged between .41 and .94 which higher than MLR with R 2 ranged between .34 and .84 (Pandiselvam, Manikantan, Sunoj, Sreejith, & Beegum, 2019).…”

“…Furthermore, ANN models indicated lower error measures (RMSE, MAE, and MRE) compared to MLR models for prediction of both final T A B L E 6 Stepwise regression analysis of final temperature (dependent variable) and vacuum cooling parameters (independent variables) for the vacuum cooling process of baby cos lettuce et al, 2019). The modeling of the extrusion process of coconut milk residue incorporated rice-corn based indicated coefficient of determination (R 2 ) of ANN ranged between .41 and .94 which higher than MLR with R 2 ranged between .34 and .84 (Pandiselvam, Manikantan, Sunoj, Sreejith, & Beegum, 2019).…”

Antioxidants, nutritional, antinutrients and functional characteristics of black turtle bean (Phaseolus vulgaris): synergistic and antagonistic interrelationship of epigeal germination periods

“…Water absorption capacity and oil absorption capacity. The WAC of the samples was determined at room temperature as described by Pandiselvam (2018) with slight modifications. One gram of the sample (W 1 ) was dispersed in 10 mL distilled water in a 15 mL pre-weighed centrifuge tube (the tube and the sample previously weighed as W 2 before adding water).…”

Modeling and Optimization of Process Parameters for Nutritional Enhancement in Enzymatic Milled Rice by Multiple Linear Regression (MLR) and Artificial Neural Network (ANN)

“…The modeling and optimization of treatment conditions for nutritional enhancement has had a problematic method, which has been analyzed as far as agriculture based products, food products, beverages, dairy products, and oil extraction industry products have pertained to different design methods to attain reasonable valuable resources [ 11 ]. Multiple linear regression (MLR) may be described while an experimental modeling design is employed as formulating, enhancing, and reducing complicated operations [ 12 , 13 ]. This method has the benefit of reducing the number of empirical tests and may be sufficient to provide a significantly acceptable outcome [ 13 ].…”