A comparative analysis of machine learning regression models of Whiteleg shrimp growth reared in eco-green aquaculture system

The success of aquaculture business of whiteleg shrimp (Litopenaeus vannamei) depends on its growth throughout the cultured period. The whiteleg shrimp growth is highly influenced by the condition of water quality in aquaculture pond. However, the study of shrimp growth reared in eco-green aquaculture system is limited. Therefore, this study aims to compare and analyze shrimp growth in relation with water quality factors in eco-green system by using machine learning regression models, such as neural network multi-layer perceptron (NN-MLP), support vector regression (SVR), decision tree, and random forest. The data was collected from 2021–2023, including average daily growth of shrimp as well as eleven water quality variables that measured both in-situ and ex-situ. The results revealed that among the five machine learning techniques employed in this study, the models generated by decision tree and random forest surpassed those of NN-MLP and SVR on the training stage. Nonetheless, in the validation or testing phase, the outcomes were inverted. Furthermore, according to the modelling results, the water quality variables that have the biggest importance value on the daily growth prediction of shrimp in the eco-green system were dissolved oxygen and nitrate concentration, as well as total vibrio.