This project was conducted to relearn old group project about lymphoma detection using CNN.
Here I made a model to classify nasopharyngeal microscopic biopsy images to one of the three classes: lymphoma, carcinoma, and benignlesion. The dataset used in this project was provided by the Faculty of Medicine, University of Padjadjaran. For privacy purposes, I will not attach the dataset for the project. I used pretrained weight from InceptionResnetV2 as the base architecture for the CNN network. The model trained achieved validation accuracy of 94% and managed to achieve 74% on test data. This model can be further improved by managing the overfitting problem.