Schedule
Course calendar
The calendar below represents the best current understanding of how the semester might unfold. Though the exact dates of some work might change (with advanced warning), the structure here can provide you with a sense of how to plan. To incorporate this calendar into your personal Macalester Google Calendar, press the +Google Calendar at the bottom.
Day-by-day schedule
The following schedule outlines a weekly plan to prepare, practice, and explore. Please note:
- Homework, checkpoints, etc are listed on the day that they’re due though you should certainly start these assignments earlier. To this end, be sure to use the calendar above for longer term planning.
- The videos are accessible on YouTube. You can download the closed captions if you’d like to save the scripts.
- Pdfs / Rmd templates for the activities are linked here and posted on Moodle.
- Religious holidays aren’t marked on the schedule. However, the instructor supports your personal observances. Please reach out to discuss accommodations around holidays.
Before class: prepare | During & after class: practice & explore | |
---|---|---|
Week 1 (8/30 - 9/3) | ||
Tues | ||
Thur | Take it easy |
1. Introduction After class: Introduce yourself on Slack Complete activity 1 / check solutions Start Homework 1 |
Week 2 (9/6 - 9/10) | ||
Tues |
Video: Binomial model Due: Checkpoint 1 Due: Review the syllabus & respond to the welcome survey |
Probability / homework workshop |
Thur | Due: Homework 1 | 2. Thinking like a Bayesian |
Week 3 (9/13 - 9/17) | ||
Tues |
Video: Beta model Due: Checkpoint 2 |
3. The Beta-Binomial model Thursday MSCS coffee breaks start this week! Next week: MSCS & Society speaker Michael Osterholm! |
Thur | Read How to think like an epidemiologist |
4. Balancing and updating (We’ll be using laptops today) |
Week 4 (9/20 - 9/24) | ||
Tues |
Video: Poisson model Video: Gamma model Due: Checkpoint 3 Due: Homework 2 |
5. Conjugate families Consider attending or presenting at eUSR! |
Thur |
Video: MCMC Video: MCMC with rstan Due: Checkpoint 4 |
6. Simulation: MCMC with rstan (We’ll be using laptops today onward) |
Week 5 (9/27 - 10/1) | ||
Tues | Nothing! Keep working on Homework 3 | 7. Posterior estimation & hypothesis testing |
Thur |
Due: Homework 3 Video: Posterior analysis Due: Checkpoint 5 |
8. Posterior prediction 9. Regression review |
Week 6 (10/4 - 10/8) | ||
Tues | Due: Checkpoint 6 |
10. Bayesian regression MSCS block party tomorrow! |
Thur |
If you didn’t already, complete exercises 1–9 from Tuesday’s activity and check solutions. Video: Evaluating a Bayesian regression model Due: Checkpoint 7 |
11. Model evaluation & comparison |
Week 7 (10/11 - 10/15) | ||
Tues |
Due: Homework 4 NOTE: The videos discuss MSE. The same ideas apply to any evaluation metric. Video: Validation Video: Cross-validation Recommended video: Bias-variance trade-off Due: Checkpoint 8 |
|
Thur |
Video: Logistic regression review Checkpoint 9 |
|
Week 8 (10/18 - 10/22) | ||
Tues | Revisit the Bias-variance trade-off video if you didn’t watch it yet | |
Thur | Nothing! | Fall break |
Week 9 (10/25 - 10/29) | ||
Tues | Nothing! Relax over break |
Capstone project overview 16: Regression & classification review |
Thur | Project checkpoint 1 |
17: Get excited about hierarchical models Capstone chat |
Week 10 (11/1 - 11/5) | ||
Tues |
Video: Grouped data Due: Checkpoint 10 Due: Homework 5 |
|
Thur |
Video: Hierarchical posterior analysis Due: Checkpoint 11 |
|
Week 11 (11/8 - 11/12) | ||
Tues | Due before class: Project Checkpoint 2 | Project time |
Thur |
Video: Hierarchical regression Due: Checkpoint 12 |
|
Week 12 (11/15 - 11/19) | ||
Tues |
Nothing due Make progress on your project |
21: More layers |
Thur | Due: Project Checkpoint 3 | Project time |
Week 13 (11/22 - 11/26) | ||
Tues | Due: Final homework | Project / Final homework time |
Thur | Thanksgiving break | |
Week 14 (11/29 - 12/3) | ||
Tues | Keep plugging away at your projects | Discuss final reports, presentations, & portfolios |
Thur | Keep plugging away at your projects |
End-of-course survey Discuss portfolios |