12 Ethics and Privacy

In this block we cover:

• What laws govern data science?
• What is Ethical data science?
• What is privacy?
• How can we design data science tools to protect it?
• What is Statistical discosure?
  • An introduction to differential privacy…
• What is interpretability in terms of machine learning algorithms?
• What is algorithmic bias? How can we mitigate it?

Lectures

• Ethics in Data Science
  • 12.1.1 Ethics in Data Science (Part 1, Overview) (21:24)
    • Slides
  • 12.1.2 Ethics in Data Science (Part 2, Disclosure) (27:09)
    • Slides
  • 12.1.3 Ethics in Data Sciencee (Part 3, Fairness and Interpretability) (28:19)
    • Slides

Worksheets:

• Worksheet 12.1 Ethics and Privacy in Data Science

References:

• Laws governing data science:
  • Human Rights Act (HRA 1998)
  • EU General Data Protection Regulation (GDPR 2018)
  • Data Protection Act 2018 (DPA 2018)
• Privacy:
  • The Algorithmic Foundations of Differential Privacy by Dwork and Roth (2014).
  • ONS policy on disclosure control.
  • Sweeney 1997. Weaving technology and policy together to maintain confi-dentiality. Journal of Law, Medicines Ethics, 25:98–110.
  • Narayanan and Shmatikov 2008. Robust de-anonymization of largesparse datasets (how to break anonymity of the netflix prize dataset). IEEE Sec. and Priv.
  • Statistical Disclosure Attacks by George Danezis.
• Interpretability and fairness:
  • Book: “Interpretable Machine Learning: A Guide for Making Black Box Models Explainable.” Christoph Molnar 2019.
  • Algorithmic Bias Tutorial by Francesco Bonchi with Slides from KDD 2016
  • Hardt, Price and Srebo Equality of Opportunity in Supervised Learning 2016 explored in https://blog.acolyer.org/2018/05/07/equality-of-opportunity-in-supervised-learning/.

Previous: Block 11.