Course Information

Space: Cyber space

Time: Slot 3, IITB time table.
Monday 10:35 to 11:30: Tuesdays 11:35 to 12:30, Thursdays : 8:30 to 9:25

Mailing List:

https://groups.google.com/g/iitb-cs689-july2021

Teaching Assistants: TBA

Contact: hariguru@cse.iitm.ac.in

Grading

• Paper presentations : 2*30=60

• Project : 30

• Mini-quizzes : 10

Syllabus

Fairness

Notions of fairness

Constrained classification

Algorithmic vs data unfairness

Explainability/Interpretability/Visualization

Visualizations of neural nets

Interpretable Machine Learning

Other Topics

Privacy (differential privacy, information theory)

Distributed/federated learning

Ethics in AI

Reference Material:

Fairness, Accountability and Transparency in ML : https://www.fatml.org/resources/relevant-scholarship

Fairness and ML textbook : https://fairmlbook.org/

Sample Explainability Papers: http://aix360.mybluemix.net/resources

Interpretable ML textbook : https://christophm.github.io/interpretable-ml-book/

Links

Week 0 and Week 1:

How big data is unfair: https://medium.com/@mrtz/how-big-data-is-unfair-9aa544d739de

Tutorial video: https://fairmlbook.org/tutorial1.html (Slides: https://mrtz.org/nips17/#/)

Classification chapter: https://fairmlbook.org/classification.html

Demo : https://research.google.com/bigpicture/attacking-discrimination-in-ml/

Week 2 to 7:

Preprocessing methods for fairness :

Zemel et al. 2013. Learning fair representations [Class/Project]

Postprocessing methods for fairness :

Hardt et al.2016. Equality of odds and opportunity [Reading/Project]

Training for fairness :

Agarwal et al. 2018. Reductions approach to fair classification [Reading/Project]

Zafar et al. 2017. Fairness constraints [Reading/Project]

Narasimhan et al. 2020a: Split-Frank-Wolfe. [Class/Project]

Narasimhan et al. 2020b: Pairwise fairness [Reading/Project]

Individual fairness :

Kearns et al. 2019. Individual fairness, Video [Class/Reading/Project]

Dwork et al. 2012. Fairness through awareness. [Reading/Project]

Hashimoto et al 2018. Fairness without deomgraphics. [Reading/Project]

Other fairness topics:

Kearns et al. 2018: Fairness Gerrymandering [Reading/Project]

Hiranandani et al. 2020 : Fair performance metric elicitation [Reading/Project]

Joseph et al. 2016: Fairness in Bandits. [Reading/Project]

Corbett et al. 2016 : Cost of fairness. [Reading/Project]

Zhang et al. 2016. Causal Framework for Removing Discrimination. [Reading]

Chiappa et al 2019. Causal Bayesian Networks. [Class]

Kusner et al. 2017: Counterfactual Fairness. [Class]

Popular science articles:

Nature

General texts on Fairness and Discrimination:

Kleinberg et al. Discrimination in the age of algorithms.

Kleinberg et al. Algorithms as discrimination detectors.

Kleinberg et al. Algorithmic Fairness

Week 8 to 10:

Zeiler and Fergus. 2013. Occlusion Map and Deconv layers. Visualizing and Understanding Convolutional Networks. https://arxiv.org/pdf/1311.2901

Simonyan et al. 2014. Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps. https://arxiv.org/abs/1312.6034

Zhou et al. 2015. Learning Deep Features for Discriminative Localization. https://arxiv.org/abs/1512.04150

Selvaraju et al. 2016. Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization. https://arxiv.org/abs/1610.02391

Desai et al. 2020. Ablation CAM. Link

Mahendran et al. 2015. Understanding Deep Image Representations by Inverting Them. https://openaccess.thecvf.com/content_cvpr_2015/papers/Mahendran_Understanding_Deep_Image_2015_CVPR_paper.pdf

Feature Visualisation : https://distill.pub/2017/feature-visualization/

Building blocks of interpretability: https://distill.pub/2018/building-blocks/

Interpretability and Explainability in AI: A Survey. https://arxiv.org/abs/1910.10045

Pruthi et al. 2018. Sparse Interpretable Neural Models. https://www.aaai.org/ocs/index.php/AAAI/AAAI18/paper/download/17433/16024

Ancona et al. 2018. Towards better understanding of gradient-based attribution methods for Deep Neural Networks

Pruthi et al. 2020. Evaluating Explanations: How much do explanations from the teacher aid students? https://arxiv.org/abs/2012.00893

Shanmugam et al. 2018. Explanations based on the missing: Towards contrastive explanations with pertinent negatives. https://arxiv.org/pdf/1802.07623.pdf

Doshi Velez et al. 2017. Towards Interpretable Machine Learning. https://arxiv.org/pdf/1702.08608.pdf