Is the training dataset balanced in terms of class distribution across the 24 unique labels?
Is the training dataset balanced in terms of class distribution across the 24 unique labels?: a task in data agent rl environment train (Harbor dataset). You have access to the following files: - sign mnist train.csv - sign mnist test.csv All of the files are located only in the '/home/user/input'…
The task
You have access to the following files: - sign_mnist_train.csv - sign_mnist_test.csv All of the files are located only in the '/home/user/input' folder without any folders inside 'input'. Do not use '/kaggle/input/' folder as it does not exist.