Project Team: Stefan Weißensteiner, Marcel Salvenmoser
- Dataset: https://www.kaggle.com/datasets/samuelcortinhas/apples-or-tomatoes-image-classification/
- Results: https://malthee.github.io/akt3/AKT3ILVT1_Weissensteiner_Salvenmoser.html
- 2 classes (apples and tomatoes)
- collect training, validation and test data
- implement some CNN
- define some experiments
- experiment with different learning rates, how does the loss behave?
- how do different CNNs perform, for example a simple CNN and MobileNetV2
- run training
- plot loss over iterations
- run evaluation on test set
- calculate confusion matrix and accuracy
- plot receiver operating characteristic [ROC] curve and calculate area under the curve [AuC]
- document your findings and your code (e.g Jupyter Notebook)
- save model, load model, run inference