<![CDATA[The texture of meatballs is one of the parameters used by the community to determine the quality of meatball products. The problem is to determine the accuracy comparison of meatball meat using four different portrait distances using the JST backpropagation method with GLCM feature extraction. Training data and test data are extracted using GLCM features, and then JST training is conducted using 17 training functions. The portrait distances used are 10 cm, 13 cm, 15 cm, and 18 cm, with 5 neurons, 10 neurons, and 20 neurons, resulting in three different architectures. For each training function, there are five experiments for each architecture. Based on the research results, it can be concluded that the traincgb training function, which uses 5 neurons, achieves fairly good recognition results at a portrait distance of 15 cm with a total accuracy of 95.5%, precision of 88%, and recall of 95%.]]>
Copyrights © 2023
