Part 2. Basic Deep Learning Practice Lab 10-1: Sigmoid 보다 ReLU가 더 좋아 Lab 10-2: Weight 초기화 잘해보자 Lab 10-3: Dropout Lab 10-4: Batch Normalization