Klasifikasi Penyakit Daun Jagung Menggunakan Model Deep Learning EfficientNetB5
Abstrak
The second most important food crop in Indonesia, after rice, is corn, which is highly susceptible to leaf diseases such as common rust, cercospora gray leaf spot, and northern leaf blight. In spite of traditional Machine Learning, in which manual feature extraction must be impeccable for optimal results, a model capable of accurate classification is required. In this investigation, the Deep Learning model EfficientNetB5 is used to classify corn leaf diseases, and the performance model between learning rate and batch size hyperparameters is compared. All models are identical to the dataset, which consists of 3,852 images divided into 4 classifications. The testing results indicate that the combination of learning rate = 0.0001 and batch size = 32 obtains the highest value compared to other models. The obtained evaluation values were 96.27 % accuracy, 90.90 % precision, 97.55% specificity, and 88.13 % sensitivity.
Identifier
- OpenAlex: https://openalex.org/W4411867853
- DOI: https://doi.org/10.56873/jpkm.v9i1.5322

