Overview#

Course Info#

  • graduate course, focused on research adjacent skills

  • topic is how to use ML safely and reliably in the context of scientific discovery or social applications

  • Classes will mostly be discussion

  • We’ll rotate leading the discussion

  • we’ll rotate note taking

Intros and Topics of Interest#

  • how to understand bias and what can be done, multiple dimension to explore

  • more about reading and writing papers

  • more skill in reading research papers

  • missing data, incomplete problems

  • HCI

  • breadth, more research

  • ML

  • eg (pain area)

  • noisy data

  • natural disaster evacuation plan

  • incomplete data

  • NLP

Overview of Course Topics#

  • COMPAS Example

  • disparate treatment/impact

  • medical

Prepare for the next class#

Prepare for Wednesday:

Model Based ML: https://mbmlbook.com/toc.html

Read: Chapter 1 & the Interlude on the ML life cycle Skim the intro to two application chapters Be prepared to compare this view of ML to how you’ve learned int (or other CS topics previously)

Read: https://web.stanford.edu/class/ee384m/Handouts/HowtoReadPaper.pdf

Be prepared to ask questions about how to prepare for presenting a paper in class

Create or make sure you can log into  GitHub Account