Which of the following statements is ( are ) true for training a neural network? (pick one or more of the following choices.) Over-fitting is caused by training using the training data for too many epochs than necessary. K-fold Cross-Validation is used to improve the accuracy of the evaluation of the neural network during training. K-fold Cross-Validation is used to speed up the training process of a neural network. Transfer learning is used to speed up the training process of a neural network