<![CDATA[Abstract Deep Learning is a subfield of Machine Learning that uses artificial neural networks with many layers to analyze and extract features from data. The Deep Learning method used in this research is the Convolutional Neural Network (CNN) method. The problem raised in this research is the classification of the quality of Cashew Seeds between good quality and poor quality. This is considered important because there has been no previous research discussing this topic. The aim of this research is to be able to classify the quality of cashew nuts based on their image using the CNN method, and see how well the CNN model that was built performs. This research uses 1000 datasets as input. The dataset consists of 2 classes, namely, 500 images of good quality cashew seeds and 500 images of poor quality cashew seeds. The classification process is carried out using several different parameters and uses the EfficientNetV2M architecture and the ResNet50 architecture. This aims to compare and determine the best results obtained from the classification system. In testing, the EfficientNetV2M architecture produces the best accuracy of 98% from a learning rate of 0.001 and epoch 60. Meanwhile, the ResNet50 architecture produces the best accuracy of 93% from a learning rate of 0.001 and epoch 20. Based on these results, it is concluded that the CNN model built has good performance in classifying quality of cashew seeds. Keywords— Deep Learning, CNN, Cashew Seeds.]]>
Copyrights © 2024
