Schedule

*This schedule is evolving and will change based on your interests and how the class is progressing.
Lec. Date Topic Slides Deadlines
1
Sept. 10
Introduction Lecture00, Lecture01
2
Sept. 17
Convolutional Neural Networks Lecture02 Indicate two prefered papers due
3
Sept. 24
Recurrent Neural Networks Lecture03 Assignment 1 is out! pdf, Code, Tutorial
4
Oct. 1
Regularization, Batch normalization Lecture04
One-Shot Learning
  • Siamese Neural Network for One-Shot Image Recognition (Adarsh, slides)
  • FaceNet: A Unified Embedding for Face Recognition and Clustering (Shubhank, slides)
5
Oct. 8
One-Shot Learning, Graph Neural Networks
  • Semi-Supervised Classification with Graph Convolutional Networks (Abdullah, slides)
  • Graph Attention Networks (Abdullah, slides)
  • Matching Networks for One Shot Learning (Omid, slides)
  • Low Data Drug Discovery with One-Shot Learning (Amirhossein, slides)
  • Learning to Remember Rare Events (Geetika, slides)
Assignment 1 due
6
Oct. 15
Domain Adaptation, Multi-task Learning, Distilling the Knowledge, Multiple Instance Learning
  • A Hybrid Instance-based Transfer Learning Method (Nkechinyere, slides)
  • Unsupervised Visual Domain Adaptation Using Subspace Alignment (Yingjun, slides)
  • Identifying beneficial task relations for multi-task learning in deep neural networks (Isabelle, slides)
  • Deep Mutual Learning (Vasileios, slides)
  • Multiple Instance Learning: A Survey of Problem Characteristics and Applications (Sedevizo, slides)
Project Proposal due. Assignment 2 is out! pdf(corrected), Code.
Oct. 22
Fall Break No class
7
Oct. 29
Zero-Shot Learning
  • A survey of zero-shot learning: Settings, methods, and applications (Michael, slides))
  • Attribute-Based Classification for Zero-Shot Visual Object Categorization (Yingjun, slides)
  • Unsupervised Domain Adaptation for Zero-Shot Learning (Nkechinyere, slides)
  • Label-activating framework for zero-shot learning (Razieh, slides)
Assignment 2 due Assignment 2 extended deadline Nov 1
8
Nov. 5
Zero-Shot Learning
  • Zero-shot learning via joint latent similarity embedding (Razieh, slides)
  • Latent Embeddings for Zero-shot Classification (Sedevizo, slides)
  • Semantic Autoencoder for Zero-Shot Learning (Vasileios, slides)
  • Relative Attributes (Mohammad, slides)
9
Nov. 12
Clustering
  • A Co-training Approach for Multi-view Spectral Clustering (Yousef, slides)
  • A survey of clustering with deep learning: From the perspective of network architecture (Adam, slides)
  • Unsupervised deep embedding for clustering analysis (Adarsh, slides)
  • Towards K-means-friendly Spaces: Simultaneous Deep Learning and Clustering (Adam, slides)
10
Nov. 19
Interpretable AI
  • Why Should I Trust You?: Explaining the Predictions of Any Classifier (Diana, slides)
  • A Unified Approach to Interpreting Model Predictions (Isabelle, slides)
  • Why Is My Classifier Discriminatory? (Geetika, slides)
  • Towards A Rigorous Science of Interpretable Machine Learning (Diana, slides)
  • Attention is not Explanation (Jacob, slides)
11
Nov. 26
Interpretable AI
  • Anchors: High-Precision Model-Agnostic Explanations (Omid, slides)
  • Explanations based on the Missing: Towards Contrastive Explanations with Pertinent Negatives (Mayank, slides)
  • Examples are not enough, learn to criticize! criticism for interpretability (Yousef, slides)
  • Interpretability beyond feature attribution: Quantitative Testing with Concept Activation Vectors (Mohammad, slides)
12
Dec. 3
Project Presentations
Project reports due on Dec. 14