UoMResearchIT/machine-learning-neural-python

Preprocessing

DrJonnyT opened this issue · 0 comments

#Print the shapes of the effusion and normal datasets and labels
print("Effusion dataset shape:", np.shape(dataset_effusion))
print("Effusion labels shape:", np.shape(label_effusion))
print("Normal datasets shape:", np.shape(dataset_normal))
print("Normal labels shape:", np.shape(label_normal))

#Print the shapes of the combined datasets
print("Combined datasets shape:", np.shape(dataset))
print("Combined labels shape:", np.shape(labels))

Also swap them round? Normal label is 0

Would be good to show the impact of preprocessing, show an image before and after

Show the image. Show the mean, min, max, stdev.
Then preprocess. Show the image again (looks the same). Show the mean, min, max, stdev (different).
Preprocessing keeps the information (hopefully) but makes the data easier for a neural network to work with