Prediction Of Milkfish Harvest Potential Based On Pond Environment To Support Smart Fishery Towards Technology 4.0

The production of fishery cultivation for simple pond commodities for three years (2017-2019) tends to continue to increase by 16.32%. Less optimal fish yields both in terms of quality and quantity are due to poor water quality management, such as unstable pH and temperature. The water quality department must have regular observations of the water quality in the ponds to be able to find out the state of the water in the ponds. For fish production results and budgets to be optimally prepared, farmers must be able to predict the yields that will be obtained in the future. However, the problem occurs because the planned production results that have been targeted are different from the production results produced. To address these issues, the author creates a website utilizing the Naive Bayes algorithm, a data mining algorithm. Many variables can affect crop production predictions, including the number of seeds planted, the size of the field, and the quality of the water, including pH and temperature. The TP value is 9, the TN is 8, the FP is 0, and the FN is 1. So that the Precision value is 1, Recall is 0.9, and F1-Score is 0.947. An accuracy number of 94.44%, which is very near to 100%, is produced by evaluating the Naive Bayes algorithm in the confusion matrix mode, suggesting that the resulting model exhibits maximum outcomes.