Machine learning vs. deep learning
Logistic regression and multilayer networks
Deep learning and optimization (e.g. weight initialization, regularization, data and batch Normalization, dropout, …)
Information Theory and Cost/Loss Function
Convolutional neural networks (e.g. onvolution and pooling, modern architectures) and object detection
Sequence modeling (e.g. long short term memory networks, memory augmented networks)
Embedding and representation learning (e.g. variational autoencoder, Word2Vec)