ECS189G is a Deep Learning upper division course we covered important topics such as:
- basic math, optimization, machine learning background knowledge
- neural network basics, error backpropagation algorithm
- auto-encoder model for data encoding and re-construction
- convolutional neural network (CNN) for computer vision
- recurrent neural network (RNN) for natural language processing
- graph neural network (GNN) for network embeddings
- graph-bert and GResNet
There were individual assignments and a group-level programming project (this project)
- Stage 1: Group formation and programming environment setup. (0%)
- Stage 2: First trial with PyTorch and build up a Multi-Layer Perceptron model to classify the instances in a provided dataset. (10%) https://docs.google.com/document/d/1xjyygpgekKi8BO9eUQqEYtdkE4ZGJWCp1JQutSmRaFw/edit?usp=sharing
- Stage 3: Object recognition from images with convolutional neural network model. (10%) https://docs.google.com/document/d/1nHUS4pkn8qVrbl28674RqYw6IB8qu5Mjn-V_TUEWqtE/edit?usp=sharing
- Stage 4: Text classification and generation with recurrent neural network models. (10%) https://docs.google.com/document/d/1Je5nkTF0nnyeTxb9uioU3LCYznZdbszN4Gjfdos1my0/edit?usp=sharing
- Stage 5: Network embedding and node classification with graph neural network models. (10%) https://docs.google.com/document/d/14BZSKBDUAZYrpaQgEhEsO2MBfe1Ey1oqa02WsmNk17A/edit?usp=sharing