Project Check-in 1
Objective:
Present 5-7 minute presentation on a Bayesian article, blog or tutorial
You may choose an academic paper, a blog post, or tutorial on a programming language or open source package.
You are welcome to choose your own or alternatively choose from any of the applications or ideas below.
Option 1: Article or Blog
Create presentation summarizing the article and blog (e.g., PowerPoint, Google Slides, Keynote, xaringan)
Include in your presentation:
First slide: What is the background? What makes it Bayesian?
2-4 Slides: Explain at least two important figures.
Last Slide: Questions or Observations
Option 2: Programming tutorial or notebook
Present a demo of the notebook (e.g., Colab, binder, GitPod, or your local computer)
Example presentation
Example Article: Vincent Warmerdam’s Blogpost on Regression per Stat School
FAQ
Can I work with someone?
Individual only. Final project you can work with someone else.
Can I pre-record my talk and submit that?
Yes! You can pre-record your talk and submit a .mp4 video recording as an alternative.
How do I present a tutorial / notebook?
Either run real time locally, on the cloud, or review static rendering (e.g., html or pdf). If you choose this option, you can simply recreate the original notebook code. What you’ll be expected is to understand and explain what the code is doing.
If I do a tutorial / notebook, do I need to use a new data set than the original tutorial or notebook?
No. You can use the original if you can explain the code/problem and run it on your own. Spend more time to create replicable reruns like cloud versions that other students can run as well (e.g., RStudio.Cloud, Colab, binder, GitPod, etc.).
I don’t understand everything in the post. I’m worried I may not know everything. Is that okay?
Absolutely! You’re not expected to know everything. Part of this exercise is learning to be confident to be comfortable with being confused.
I recommend keeping your slides to only images or very short bullet points. Many students have the urge to do the opposite – add as much to the slide to make up for worries of not knowing everything. Just state items you don’t understand, especially in your conclusion.
Suggested References
Sciences
| Topic | Reference | Type | Language |
|---|---|---|---|
| Bayesian Model of Planetary Motion | Bayesian Model of Planetary Motion | R | |
| Exoplanet Detection | Intro to gaussian processes in Stan: Finding exoplanets | R | |
| Covid Vaccine Efficacy | Bayesian Statistics of Efficacy of the Pfizer-BioNTech COVID-19 Vaccine — part I | ||
| Bayesian Cognitive Modeling | A Beginner’s Guide to Bayesian Cognitive Modeling | Turing |
Time or Space
| Topic | Reference | Type | Language |
|---|---|---|---|
| Games (Battleship) | Playing Battleship with Bayesian Search Theory, Thompson Sampling, and Approximate Bayesian Computation | Python | |
| Intro to Gaussian Processes | Visual exploration of Gaussian processes | ||
| Bayesian Time Series | Uber’s Time Series Inference and Forecasting package | Python |
Financial / Political
| Topic | Reference | Type | Language |
|---|---|---|---|
| Bayesian Portfolio Analysis | Bayesian Portfolio Optimisation: Introducing the Black-Litterman Model | Python | |
| Visualizing Bayesian Unemployment Forecasts | Uncertainty examples with US unemployment data | R | |
| Election Forecasting | Economist 2020 Election Forecasting | ||
| Election Forecasting | An Updated Dynamic Bayesian Forecasting Model for the US Presidential Election |
Machine Learning
| Topic | Reference | Type | Language |
|---|---|---|---|
| Fairness in ML | Great Potential of Priors: Common Sence Reduced to Priors | Python | |
| Fairness in ML | How You Can Add Fairness Constraints to Models Using Priors | Python |
Marketing / Consumer Behavior
| Topic | Reference | Type | Language |
|---|---|---|---|
| A / B testing | R package bayesAB tutorial | R | |
| Bayesian Optimization | Lyft Bayesian Optimization | ||
| Bayesian Optimization | Apple’s Interpretable Adaptive Optimization | ||
| User Ratings | A Bayesian Model of Lego Set Ratings | Python | |
| Recommender & Multiarmed Bandits | A penguin fish-recommender systems using multi-armed bandits pt. 1 | Julia |
Human-Computer Interaction and Visualization
| Topic | Reference | Type | Language |
|---|---|---|---|
| Uncertainty Visualizations in Mobile Apps | Uncertainty displays using quantile dotplots or CDFs improve transit decision-making (Fernandes et al., 2018) | R | |
| Bayesian Methodology in HCI | Why Bayesian Statistics Better Fit the Culture and Incentives of HCI (Kay et al., 2016) | R | |
| Bayesian Data Visualizations | Evaluating Bayesian Model Visualisations (Stein & Williamson, 2022) |
Natural language programming
| Topic | Reference | Type | Language |
|---|---|---|---|
| Original paper on LDA (topic modeling) | Latent dirichlet allocation (Blei, Ng, Jordan, 2003) | ||
| Introduction to Topic Modeling | Probabilistic Topic Models | ||
| Topic Modeling in Python | OCTIS python package | Python | |
| Predicting Federalist Paper Authorship | Naive Bayes | R |
Sports
| Topic | Reference | Type | Language |
|---|---|---|---|
| Baseball statistics | Understanding Bayesian A/B testing (using baseball statistics) | R | |
| Baseball statistics | Understanding credible intervals (using baseball statistics) | R | |
| Baseball statistics | Hierarchical Partial Pooling of Batting Averages | Python (PyMC3) | |
| Soccer | A Bayesian Approach to In-Game Win Probaiblity | R |
Probabilistic programming
| Topic | Reference | Type | Language |
|---|---|---|---|
| Any Stan Code Examples | Stan website | Many | |
| Any Turing.jl Code Tutorials | Turing.jl website | Julia | |
| Any PyMC3 Code Examples | PyMC website | Python |