HUDM 5127
Spring 2022

Project Guidelines

Choose a topic from the following list:
1. Logistic regression
2. Poisson regression
3. Nominal regression
4. Ordinal regression
5. Survival analysis
6. Mixtures
7. Multilevel models
8. Missing data

• Projects are individual.
• Choose a topic from the list. Further down you go on the list, the more complex
the topic is.
• Students who choose “easier” topics, will be expected to go the extra mile and
make their reports outstandingly good.
• Find a suitable dataset for your topic.
• Analyze with Rstan ulam (no quadratic approximation map function will be
accepted as main result!).
• Your project should contain the following:

Ø Introduction to the topic you chose (what it is used for, how it works,
limitations, assumptions, and so on);
Ø Description of the data you will use : variables, what you want to achieve
with it (research question), and how it will be done;
Ø Prior choice: how and why you decided to use such priors. You will also
be expected to try at least two different priors and compare results.
Ø Analysis results (Rstan output, interpretation and conclusions):
convergence diagnostics, tuning the chain(s) if necessary;
Ø Summary of findings, comparisons, at least one graph of the final fitted
model and what the results imply about the data and research question.


欢迎咨询51作业君