Application of artificial neural network‐genetic algorithm and response surface methodology for optimization of ultrasound‐assisted extraction of phenolic compounds from cashew apple bagasse

Abstract

This study explores the effect of ultra-sonication on the extraction of phenolic compounds of cashew apple bagasse (CAB). The extraction was conducted at different combinations of treatment time (5-15 min), ultrasound amplitude (30-60%), and bagasse to solvent ratio (30-50 g/ml) for maximum yield of total phenolic content (TPC), total tannin content (TTC), and β-carotene content. Both response surface methodology (RSM) and artificial neural network (ANN) were used to model and optimize the process parameters. The optimum conditions achieved using RSM were: treatment time of 12 min, ultrasound amplitude of 55%, and bagasse to solvent ratio of 45 g/ml. In contrast, the optimum conditions proposed by ANN were: treatment time of 15 min, 60% amplitude, and 50 g/ml bagasse to solvent ratio. The TPC, TTC, and β-carotene content was determined under RSM optimized conditions as 29.04 mg GAE/g, 26.70 μg tannic acid/g, and 18.12 μg/g, respectively and under ANN optimized conditions as 30.61 mg GAE/g, 30.80 μg tannic acid/g, and 19.50 μg/ g, respectively. All the statistical parameters showed the lowest value in the ANN model than the RSM model with the highest coefficient of determination (R 2 ). The measured value of the TPC, TTC, and β-carotene content at optimized conditions were in agreement with the predicted value obtained from the above models. Besides these, the ultra-sonication significantly increased the TPC, TTC, and β-carotene content while comparing with conventional solvent extraction. Hence, this method could be an effective green method over conventional technologies to extract phenolic compounds from CAB.

Optimization of ultrasound‐assisted extraction of ascorbic acid, protein and total antioxidants from cashew apple bagasse using artificial neural network‐genetic algorithm and response surface methodology

“…For the development of the ANN model, the input and output data points were coded. The data set of the input neuron was coded between −1 to +1, and the dataset of the output neuron was coded between 0 to +1 (Patra et al., 2021a; Pradhan et al., 2021; Sonawane et al., 2020). The hidden layer and its number of neurons were decided using the trial and error technique to achieve the best optimum condition (minimum mean square error [ MSE ] and maximum R 2 ) (Raj & Dash, 2020; Sonawane et al., 2020).…”

“…The bagasse was separated and dried in a hot air oven at 50°C for 24 hr. Then the dried bagasse was sieved using a 60-mesh sieve, and then it was stored in moisture-proof polythene bags at 4°C for further analysis (Patra et al, 2021a(Patra et al, , 2021b.…”

Pressure‐driven crossflow microfiltration coupled with centrifugation for tannin reduction and clarification of cashew apple juice: Modeling of permeate flux decline and optimization of process parameters

“…Only a single hidden layer was used to avoid the generalization error and issues related to over‐fitting (Abdullah et al, 2021). In addition, the number of hidden layer neurons was selected based on the trial and error (heuristic) method with maximum R 2 and least mean squared error (MSE) values for the output data (Patra et al, 2021b). The tansig (hyperbolic tangent sigmoid) and purelin (linear transfer function) were used as the transfer function to calculate the hidden and output layer, respectively.…”

“…Ascorbic acid content, expressed in mg/100 ml, was determined by visual titration using 2,6‐dichlorophenol indophenol (Patra et al, 2021a). Total tannin content, expressed in mg/100 ml, was determined using (Abdullah et al, 2020; Patra et al, 2021b). Total phenolic content, expressed in mg Gallic acid equivalent per 100 ml (mg GAE/100 ml), was determined using the Folin–Ciocaltaeu method (Ghosh et al, 2018; Patra et al, 2021b).…”

“…The tannin is a water-soluble secondary metabolite present in different plants. It is a phenolic compound that is also present in various fruit juices, including CA juice (Patra et al, 2021b). The tannin forms complexes with protein (especially salivary proteins such as proline, histatins, and glycoprotein) and increases the haziness and turbidity of the juice.…”