Klasifikasi Penyakit Daun Cabai Menggunakan Model MobileNet

Chilli is the most popular and strategic commodity to reduce inflation in Indonesia. Chili consumption in Indonesia has continued to increase from 2019 to 2023 but chili production in Indonesia has decreased from 2021 to 2023 as well as the chili harvest area which has decreased from 2020 - 2023 which has resulted in obstacles in meeting the needs of chili consumption in Indonesia. The problem that occurs is due to chilli farmers who experience crop failure due to disease in the chilli plant.With these problems, there needs to be a solution to overcome these problems, therefore an application is made that can classify diseases on chili leaves so that chili farmers or the general public who want to plant chilies can detect diseases on chili leaves early. The model made using the CRISP-DM method by applying the concept of supervised learning with the Convolutional Neural Network (CNN) algorithm and MobileNet architecture.From the results of the research that has been done, the model results have good performance and do not show signs of severe overfitting with the results of the training accuracy model evaluation of 94.18%, validation accuracy of 83.66%, training loss of 16% and validation loss of 45%. The confusion matrix results have an accuracy of 84.84% of the total testing data of 132 data.