Responsible machine learning in Python

This lesson explores key topics on the responsible application of machine learning. The lesson is presented as a series of case studies that illustrate real world examples. Sections cover a broad range of topics, including reproducibility, bias, and interpretability. Broadly the topics are ordered chronologically, appearing as they would when thinking through a research study.

Prerequisites

You need to understand the basics of Python before tackling this lesson. The lesson sometimes references Jupyter Notebook although you can use any Python interpreter mentioned in the Setup.

Getting Started

To get started, follow the directions on the “Setup” page to download data and install a Python interpreter.

Schedule

	Setup	Download files required for the lesson
00:00	1. Introduction	What do we mean by responsible machine learning? What types of harm may result from development and deployment of machine learning models? What steps are being taken to mitigate the risks of harm?
00:30	2. Tasks	Which tasks are appropriate for machine learning? What are the principles of ethical machine learning?
01:00	3. Data	How does data influence machine learning? How can we better document data?
01:30	4. Fairness	What do we mean by fairness and bias? What are some examples of biased models?
02:00	5. Dataset shift	What is dataset shift? What are examples of dataset shift? What are the implications of dataset shift?
02:30	6. Explainability	What is explainability? Is explainability necessary?
03:00	7. Attacks	How can models be intentionally mislead?
03:30	Finish

The actual schedule may vary slightly depending on the topics and exercises chosen by the instructor.