Department of Agricultural Engineering, Agricultural Technology Faculty, Universitas Brawijaya, Veteran St., Malang, Indonesia, 65145

Email: istifaryogi_p@student.ub.ac.id

The objective of this research is to classify the quality of dried clove flowers using the deep learning method with the Convolutional Neural Network (CNN) algorithm and perform sensitivity analysis or CNN hyperparameters to obtain the best model form in the clove quality classification process. This research uses clove raw material which has been classified according to SNI 3392-1994 by PT. Perkebunan Nusantara XII Pancusari Plantation, Malang Regency. The number of samples used is 1600 dried clove flower image data which is divided into 4 qualities. Each clove quality has 225 training data, 75 validation data, and 100 test data. There are several stages in this research, the first is making the CNN model architecture the first model. The results were obtained from the reading accuracy of 65.25%. The second stage is to analyze the CNN sensitivity or CNN hyperparameter on the first model. By taking the model that has the best value at each stage of the CNN hyperparameter to be used in the next stage. The final result after the CNN hyperparameter was carried out, the test data reading accuracy value was 84.25%.