Course Schedule

Week 1: Jan. 20

• Topics: Course overview
• Course Materials: slides

Week 2: Jan. 25, 27

• Topics: Introduction to interpretability, Interpretable generalized additive models (GAMs)
• Course Materials: slides_1, slides_2

Week 3: Feb. 1, 3

• Topics: Introduction to neural networks
• Course Materials: slides

Week 4: Feb. 8, 10

• Topics: Post-hoc explanations for black-box models: perturbation-based methods
• Course Materials: slides

Week 5: Feb. 15, 17

• Topics: Post-hoc explanations for black-box models: gradient/attention-based methods
• Course Materials: slides

Week 6: Feb. 22, 24

• Topics: Post-hoc explanations for black-box models: beyond feature-level
• Course Materials: slides

Week 7: Mar. 1, 3

• Topics: Improving neural network intrinsic interpretability
• Course Materials: slides

Week 8: Spring Recess

Week 9: Mar. 15, 17

• Topics: Building interpretable neural network models
• Course Materials: slides

Week 10: Mar. 22, 24

• Topics: Rationalized neural networks
• Course Materials: slides

Week 11: Mar. 29, 31

• Topics: Interpretation and human understanding
• Course Materials: slides

Week 12: Apr. 5, 7

• Topics: Robust interpretations
• Course Materials: slides

Week 13: Apr. 12, 14

• Topics: Interpretations for improving model performance, robustness, fairness
• Course Materials: slides

Week 14: Apr. 19, 21

• Paper Discussion

Week 15: Apr. 26, 28

• Paper Discussion

Week 16: May 3

• Final project presentation