Artificial neural network and random forest regression models for modelling fatty acid and tocopherol content in oil of winter rapeseed

Abstract

With the aid of models used in artificial intelligence, a wide range of data can be processed quickly with high accuracy. The quality of rapeseed oil from 40 genotypes cultivated during four consecutive years was analysed. Two machine learning techniques (artificial neural network - ANN, and random forest regression - RFR) were applied for the modelling of fatty acids content (C16:0; C18:0; C18:1; C18:2; C18:3 and C22:1), alpha-tocopherol, gamma-tocopherol and total tocopherols, according to the data of production year and winter rapeseed genotype. The developed models exerted high-quality anticipation features, showing high r2 during the training cycle. The best fit between the modelled and measured traits for ANN model was observed for erucic acid content. RFR modelling for all fatty acids was more effective than ANN model, with the highest precision for palmitic, stearic, and oleic fatty acids (r2>0.9). This study emphasized the possibility of using ANN and RFR models to model winter rapeseed quality traits.