what is the difference between testing and validation data??
Image Pipelines
Both are kind of same things.
There is a very popular way of dividing the data.
80% in training set
10% in validation set
10% in test set.
Train for 80% of data, check the accuracy on 10% validation data, and you do this for multiple times for different different models try to improve the validation accuracy… Once you are satisfied on val accuracy,
Do a final testing on 10% test set, and that would be your final accuracy on unseen data
Thanks 