During mini-batch training of a neural network for a classification problem, a Data Scientist notices that training accuracy oscillates.What is the MOST likely cause of this issue?
A. The class distribution in the dataset is imbalanced.
B. Dataset shuffling is disabled.
C. The batch size is too big.
D. The learning rate is very high.