-
Notifications
You must be signed in to change notification settings - Fork 0
/
Schedule.qmd
72 lines (53 loc) · 8.5 KB
/
Schedule.qmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
---
title: "Schedule: Rethinking Seminar - Winter 2023"
author: "Jon Nations"
format:
html:
toc: true
toc-depth: 5
editor: visual
date: last-modified
---
#### **Week 1**
- Intro to Course, R, and RStudio.
#### **Week 2**
- Rethinking: Chapters 1, 2, & 3
- Lectures: [Video 1: The Golem of Prague](https://www.youtube.com/watch?v=cclUd_HoRlo&list=PLDcUM9US4XdMROZ57-OIRtIK0aOynbgZN&index=2) & [Video 2: Bayesian Inference](https://www.youtube.com/watch?v=guTdrfycW2Q&list=PLDcUM9US4XdMROZ57-OIRtIK0aOynbgZN&index=3)
- Rethinking in brms: Chapters [1](https://bookdown.org/content/4857/the-golem-of-prague.html), [2](https://bookdown.org/content/4857/small-worlds-and-large-worlds.html), and [3](https://bookdown.org/content/4857/sampling-the-imaginary.html). In addition to the book content, these chapters provides a good intro the tidyverse and brms. Rethinking uses Quadratic Approximation for the first few chapters (`rethinking::quap`function), while brms only does full Bayesian sampling. Sections [2.4.3](https://bookdown.org/content/4857/small-worlds-and-large-worlds.html#grid-approximation.) to [2.4.5](https://bookdown.org/content/4857/small-worlds-and-large-worlds.html#markov-chain-monte-carlo.) demonstrates the differences, and section [3.4](https://bookdown.org/content/4857/sampling-the-imaginary.html#summary-lets-practice-with-brms) shows how it all works in brms.
#### **Week 3**
*Starting Here* I changed the videos from 2022 lectures to 2019 Lectures<br>
- Rethinking: Chapter 4
- Lectures: [Video 3: Geocentric Models](https://www.youtube.com/watch?v=zYYBtxHWE0A), & [Video 4: Categories, Curves, and Splines](https://www.youtube.com/watch?v=QiHKdvAbYII&list=PLDcUM9US4XdMROZ57-OIRtIK0aOynbgZN&index=5)
- Rethinking in brms: [Chapter 4](https://bookdown.org/content/4857/geocentric-models.html#curves-from-lines)
- **Note** - 4.1 - 4.4 are the key sections. Regardless of the book cover, time series and splines are not the focus of this seminar. If you do deal with lots of time data in your discipline, then spend some extra time on it.
#### **Week 4**
- Rethinking: Chapters 5 - Intro to Multiple Regression
- Lectures: [Video 5](https://www.youtube.com/watch?v=e0tO64mtYMU&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=5), [Video 6](https://www.youtube.com/watch?v=l_7yIUqWBmE&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=6)
- Rethinking in brms: [Chapter 5](https://bookdown.org/content/4857/the-many-variables-the-spurious-waffles.html)
- **Note** - Jon will do a quick walk through of Chapter 6.1 because it's an important issue that, I swear, some reviewer will bring up to each of you at some point! the rest of Chapter 6 focuses on scientific models and directed acyclic graphs. This is really interesting stuff, and worth revisiting on your own when thinking about variable selection, and forming your models. You can preview Jon's 6.1 page [here](https://htmlpreview.github.io/?https://github.com/jonnations/Rethinking_Seminar_Winter_2023/blob/main/Multicollinearity.html) or look at the code [here](https://github.com/jonnations/Rethinking_Seminar_Winter_2023/blob/main/Multicollinearity.qmd) if you're curious how it's put together.
#### **Week 5**
- Rethinking:Chapter 8
- Lectures:[Video 9](https://www.youtube.com/watch?v=QhHfo6-Bx8o&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=9).
- Rethinking in brms: [Chapter 8](https://bookdown.org/content/4857/conditional-manatees.html) (*and it is worth going through [Chapter 7](https://bookdown.org/content/4857/ulysses-compass.html) and [Chapter 9](https://bookdown.org/content/4857/markov-chain-monte-carlo.html)*)
- **Note**: Here we start really jumping around. Chapter 8, on interactions, is a really key part of this course and what we will focus lecture time on. Chapter 7 discusses model predictive accuracy and model comparison. ***These are critical topics to Bayesian multilevel modeling*** but are very complex and time consuming to grasp, especially this early on in the topic. The same goes for MCMC machinery (Ch 9). Watch the videos, [but most of our model comparison practice will be integrated into the rest of the course]{.underline}. *As you move forward in your projects, I recommend revisiting Chapter 7 frequently!*
- Also, DAGs will be used in most of the models in the book from now on. This is a very useful thing to learn, but it takes a bit of practice. In addition to the book, [HERE](https://github.com/jonnations/Rethinking_Seminar_Winter_2023/blob/main/Cinelli%20et%20al.%20-%202022%20-%20A%20Crash%20Course%20in%20Good%20and%20Bad%20Controls.pdf) is a very useful paper on how to use DAGs with some good examples near the end.
#### **Week 6**
- Rethinking: Read/Skim Chapter 10, Focus on Chapter 11, and glance over Chapter 12, mostly just to see what GLMs are there so you can come back to it later.
- Lectures: [Second Half of Video 11](https://youtu.be/-4y4X8ELcEM?list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&t=1194), & [Video 12](https://www.youtube.com/watch?v=hRJtKCIDTwc&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=12). If you find out that you will be using some generalized linear model for your project, go through [Video 13](https://www.youtube.com/watch?v=p7g-CgGCS34&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=13) and [Video 14](https://www.youtube.com/watch?v=zA3Jxv8LOrA&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=14) and find the relevant portions and watch them. The order should be similar to that of the text book.
- Rethinking in brms: Chapters [10](https://bookdown.org/content/4857/big-entropy-and-the-generalized-linear-model.html), [11](https://bookdown.org/content/4857/god-spiked-the-integers.html). [Chapter 12](https://bookdown.org/content/4857/monsters-and-mixtures.html) is **indispensable** if you are using certain GLMs for your project.
- **Note**: Introduction to Generalized Linear Models! The types of models discussed may or may not be important to your work. For example, I use Bernoulli (logistic) and Ordered Categorical (ordinal) models all the time, but I never use Poisson models (count data). The most important things to know: 1) what are the different types of response variable distributions, and when to use them, and 2) what a **link function** is.
#### **Week 7**
- Rethinking: Chapter 13
- Lectures: [Video 15](https://www.youtube.com/watch?v=AALYPv5xSos&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=15), & [Video 16](https://www.youtube.com/watch?v=ZG3Oe35R5sY&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=16)
- Rethinking in brms: [Chapter 13](https://bookdown.org/content/4857/models-with-memory.html)
- **Note**: The big one! Multilevel models! This week will focus on the theory behind multilevel modeling, explain why you should (almost) always use them, and introduce "varying intercepts" models.
#### **Week 8**
- Rethinking: Chapter 14.1 - 14.4
- Lectures: [Video 17](https://www.youtube.com/watch?v=yfXpjmWgyXU&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=17), with [Video 18](https://www.youtube.com/watch?v=e5cgiAGBKzI&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=18) and [Video 19](https://www.youtube.com/watch?v=pwMRbt2CbSU&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=19) being optional (but it could be important to evolutionary biologists or geographers! see notes below)
- Rethinking in brms: Chapters [14.1 - 14.3](https://bookdown.org/content/4857/adventures-in-covariance.html#varying-slopes-by-construction) with 14.4 and 14.5 optional.
- **Note**: Introduction to varying slopes models. Chapters 14.1-14.3 are very critical to learn. Chapter 14.4 is a case study on social networks (any anthropologists here?), and 14.5 is a study on 2 flavors of Gaussian processes, including geospatial modeling (correlations due to geographic proximity) and phylogenetic regression (correlations due to evolutionary proximity). If you have interest in any of these things, then definitely watch Videos 15 or 16, and read the book chapters (14.4 & 14.5) carefully.
#### **Week 9**
- Rethinking: Chapter 15.1-15.2, and 17
- Lectures: [Video 20](https://www.youtube.com/watch?v=UgLF0aLk85s&list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI&index=20)
- Rethinking in brms: Chapters [15.1 and 15.2](https://bookdown.org/content/4857/missing-data-and-other-opportunities.html)
- **Note**: Wrapping up with measurement error and missing data. 15.1 and 15.2 are methods that will likely serve everyone is this course. Also *HIGHLY RECCOMMENDED* is this [brms vignette on missing data](https://cran.r-project.org/web/packages/brms/vignettes/brms_missings.html). Chapter 17 is a really nice summary of what to think about going forward.