- M123BSY1566-Daffa Maulana Muhammad
- M183BSX1278-Chindy Arendika Putri
- M183BSY1720-Akhdan Ferdiansyah Ramadhan
In our project, we collect datasets from Kaggle, a renowned platform for machine learning datasets. This particular dataset is meticulously categorized into two distinct classes: normal eyes and cataract eyes. To ensure the robustness of our model, we adopted a widely accepted split, allocating 80% of the dataset for training purposes and reserving the remaining 20% for testing. This division allows us to train our model on a substantial portion of the data, enabling it to learn intricate patterns.
This is the Cataract Datasets.
This is the model architecture that we use for training the data. We are using a Convolutional Neural Network model with One Input Layer, Six Hidden Layer, and One Output Layer.
From the model that we have trained, we get this Training Accuracy, validation accuracy, Training loss, and validation loss. This is the graphic of the training result from 20 epochs.