Evaluate given dataset to classify skin cancer images as benign or malignant.
To run :
python main.py
•main.py o This is file that you run to run training and testing o There is an args parser for the input user to setup specific arguments like what model, loss, optimizer, ect to use
•dataloader.py o custom data set class for controlling preprocessing steps o we have specific preprocessing used in the training that does not need to be used on test examples
•model_selection.py o SkinCancerModel is our nn model class, choosing the backbone of the solver is mapped out here o train_model and evaluate_model functions are here and called in main
•visualization.py o visualization code is for plots and image examples are held here and called in main o some work is done in other files to make sure the needed information is passed to these functions
•run_tests.sh o file that would be used to control running many main.py and arguments so that a person isn’t writing every command. Just a way to easily have training and testing continue automatically when setup