My I-Pond : Water Quality Monitoring with IoT and Machine Learning to Reduce Pond Cultivation Failure for Farmers

This research is focused on the application of IoT and machine learning to monitor water quality in fish ponds, poor water quality, and traditional pond processing are some of the causes of fish harvest failure. Based on this phenomenon, a new collaboration is proposed to predict water quality using the Internet of Things. This tool uses three sensors, namely turbidity, temperature, and pH sensors. The data generated from the three sensors will be processed using machine learning to train the model in predicting water quality in fish ponds, the classification of water quality is divided into two types, namely Normal, Turbid, and Very Turbid. The machine learning models used are Random Forest Classifier, XGBoost, Support Vector Machine (SVM), and Logistic Regression. Based on the research conducted, the Random Forest model produces highest performance compared to the other three models, with 96% accuracy and 96% precision, 96% recall, and 96% f1- score.