Active machine learning (ML) models have started replacing the conventional methods of crops health management and farm productivity with predictive analytics. The aim of this research paper is to anticipate the health of coconut trees, an important tropical crop, with the help of ML applications. These coconut palms are strong in economic value but also ecologically significant, and they are vulnerable to an assortment of diseases and pests which, if they are allowed to flourish, may substantially reduce yield. Conventional methods of assessing coconut tree health require observations made in the field and delayed diagnostic results, which are both labor intensive. We note that coconut tree health issues have been addressed using advanced ML models for early detection and prediction in this paper. Several ML algorithms are analyzed in the study for data from several sources like satellite imagery, drone based sensors, and field data, including Convolutional Neural Networks (CNNs), Random Forest and Support Vector Machines (SVMs). With integration of these data sources, ML models can find patterns, anomalies in health problems. The paper also describes the modeling of predictive models which can project potential outbreak of the disease and pest infestation using historical data and real time observations. Various ML models are proven to have high accuracy in detecting early signs of disease and stress in coconut trees. The use of remote sensing data in conjunction with ML algorithm results in tremendous increase in predictive capability that facilitates timely interventions and directed management strategies. This paper presents a case studies on the implementation of ML models to improve coconut tree health management with the highlight of the practical benefits and challenges of the technologies involved.
Goswami et al. (Fri,) studied this question.