Optimizing Pleurotus cornucopiae var. citrinopileatus Mushroom Extraction Conditions Using Moth-Flame Optimization Algorithm

Abstract

The aim of this study is to optimize the extraction conditions of Pleurotus cornucopiae using artificial intelligence methods. For this purpose, the data of antioxidant activities of mushrooms extracted at 0, 30, 60, 90 % ethanol ratio, 1, 2 and 4 mg/mL extract concentration and 1,2, 3, 4.2 and 6 pH conditions were obtained from an previous experimental study. The extraction conditions were modelled using artificial neural networks and optimized using Moth-Flame Optimization algorithm. In order to obtain the best prediction model, different numbers of hidden neurons were tried and the optimal number of hidden neurons was found to be 5. The mean of squares of error and mean absolute percent error of this model were found to be 1.79 and 3.24%, respectively, for the all data set. After the optimization process, the maximum antioxidant activity was found to be 56.76%, and the optimum extraction parameters were determined as 66.34% ethanol ratio, 4 mg/mL extract concentration and 2.36 pH to obtain this result. This study revealed that the use of artificial neural networks and Moth-Flame Optimization Algorithm integration provides time, labor and cost efficiency in the optimization of extraction conditions.