Klasifikasi Kualitas Terung Ungu dengan Menggunakan Metode Convolutional Neural Network

Abstract

Eggplant is an annual plant in temperate climates and tropical areas and is one of the fruit-shaped vegetables that is popular in Indonesia. With a very large number of harvests and classification that still uses manual methods of eating, another alternative is needed using Machine Learning. Convolutional Neural Network (CNN) itself is a machine learning method that is known to have a high level of accuracy. CNN classification can use a Confusion Matrix to label the class of objects detected. Based on research results based on the Confusion Matrix, it was found that the True Positive class was 52, True Negative was 54, False Positive was 6, and False Negative was 8. Then we obtained an average percentage of accuracy of 88.33%, which means detecting the quality of purple eggplant using the Convolutional method The Neural Network in this research is quite high.