Unified GIS-based machine learning method for effective forecasting of disease propagation and resistance in aquatic farming

Abstract

Integrating Geographic Information Systems (GIS) with machine learning offers a powerful tool for managing disease risks in aquatic farming. This study proposes a unified framework that leverages spatial and predictive analytics to enhance disease propagation and resistance forecasting in aquaculture environments. Traditional methods for disease prediction often rely on isolated environmental data or manual observation, which lack spatial context and fail to provide early warnings. These limitations reduce the efficiency of response strategies and increase economic losses. The proposed framework integrates GIS-based spatial mapping with supervised machine learning models trained on environmental variables, water quality indicators, and historical disease data to overcome these challenges. This combination enables accurate spatial-temporal predictions of disease outbreaks and resistance patterns. The model is applied as a decision-support system for aquaculture managers, providing interactive risk maps and predictive insights for targeted interventions and optimized farm management. Findings reveal that the unified method significantly outperforms conventional prediction accuracy and response-time approaches. It allows for more efficient resource allocation, proactive health monitoring, and improved sustainability in aquaculture practices.