Implementasi Metode Naive Bayes pada Sistem Pakar Diagnosis Penyakit Kutu Ikan Gurami (Argunus Indicus)

Abstract

Gouramy is a freshwater fish that is widely cultivated by breeders and many experience death. Many deaths are caused by various diseases such as fungi and fish lice. The presence of disease in gouramy causes losses due to the number of deaths and can reduce quality such as freshness, color, and body defects which can affect the selling price of fish or economic value. Gouramy mortality data can reach up to 50%-100%. To reduce losses due to high mortality, an expert is needed to diagnose the disease. But the fact is that not all gouramy farmers understand how to diagnose fish, therefore an expert system is needed that can be used to help farmers to diagnose fish lice disease based on symptoms. The results of the system evaluation using 20 (twenty) fish symptom data obtained from carp breeders in 2021 which were compared with expert beliefs calculated using the confusion matrix table, the accuracy values were 94.2%, precision 95%, sensitivity 95% and specivity 93.3%. The evaluation results prove that Naïve Bayes has succeeded in providing good diagnostic results, so that the developed system can be used by fish farmers in diagnosing gouramy disease. Keywords: Gourami; Diagnosis; Fish Fleas; Naive Bayes; Expert system