Multivariate Analysis of Water Quality Parameters for Sustainable Prawn Farming

Abstract

Objective: A robust framework for enhancing prawn aquaculture must be established by integrating multivariate analysis and machine learning models to evaluate and predict essential water quality parameters effectively.
Methods: The empirical water quality data taken from prawn aquaculture ponds was dimensionally reduced using Principal Component Analysis (PCA). The Successive Projections Algorithm (SPA) was used to identify key parameters, while machine learning models (XGBoost, Random Forest, and SVM) were employed to build prediction models. The models were tested for accuracy, precision, and computational efficiency.
Results: Using principal component analysis (PCA) and principal component spectral analysis (SPA), the research study was able to proficiently analyse water quality data, identify significant elements that affect prawn health, and construct and test prediction models (XGBoost, Random Forest, and SVM) that produced accurate forecasts of vital water quality indicators.
Conclusion: The integrated framework, which combines multivariate analysis and machine learning, is critical for optimising shrimp aquaculture practices. Its proactive management, continuous real-time monitoring, and precise water quality predictions enhance shrimp survival rates, mitigate disease outbreaks, and increase the sustainability of shrimp aquaculture operations.