This repository contains two folders, each one containing the notebooks used to execute the tasks discussed in the thesis. In each folder, there is a notebook for the Convolutional from Scratch approach and another for the transfer learning approach. However, there are two versions of each notebook:
- A first version containing all the outputs, with the results of the training process and the plots. You can play interactively with each plot and see the exact numbers.
- The second version of the notebook does not contain any output. Thus, Its name is the same as the last one adding -NoOutputs. Clearing all the outputs makes the code more readable.
For opening the notebooks, you can download them and open them in your personal computer, or more easily, open the file in GitHub. However, it only opens a rendered version of the notebook. Then, you should click on the link provided by Github which will redirect you to a jupyter notebook viewer, as in the picture below.
