Due Tuesday, November 3: Models, Part 2 — Uncertainty and Public Decisions
Models help us explain and predict, but their assumptions and limits matter when predictions inform public decisions. We will connect distributions and regression to the difficulty of modeling COVID-19 and making decisions with incomplete, changing evidence.
Read the COVID materials together: how do they explain uncertainty, which outcomes do they prioritize, and what happens when models disagree? How much uncertainty should officials communicate? How should institutions respond to disagreement while maintaining public trust?
1. The Model Thinker: chapters 5–7 and 29
- Scott E. Page, The Model Thinker, chapters 5–7 (pp. 59–93): normal distributions, power laws, and linear models. Skip chapter 4 for now.
- The Model Thinker, chapter 29: Opioids, COVID-19 and Inequality.
I recommend buying the book; the scanned excerpts are a backup.
2. Understanding Machine Learning, chapter 2
Complete chapter 2: Machine Learning Models of DataCamp’s Understanding Machine Learning, postponed from October 29.
Watch Visualizing a Decision Tree — Google Developers (6:41).
3. COVID modeling and pandemic policy
- Why It’s So Freaking Hard to Make a Good COVID-19 Model — Koerth et al., FiveThirtyEight (March 31, 2020).
- Stanford Pandemic Policy Opening Remarks — watch through 9:00.
- Stanford Roundtable on Pandemic Policy — watch through 51:45.
These are historical materials: consider what was known at the time and distinguish predictions, policy choices, and later assessments.
Mini-Project 2 is due Friday, November 6, at 11:59 p.m. Eastern.