Hi all!
I am implementing a rank ordered logit model using PyStan for a client. The model is described in this paper http://www.glicko.net/research/multicompetitor.pdf
I have made a lot of progress on this model but this is my first time using Stan and I have reached the limits of my knowledge. I am looking for an expert who can confirm my current implementation as well as help with the Stochastic process that happens between time periods.
I am happy to provide more details. Please reach out publicly here, or privately by email jdegrootlutzner [at] gmail.com
I think @bgoodri may have created an example for his course that uses a model similar to this for driver skill in race cars or something to that effect.
1 Like
Hi, @jdegrootlutzner. Feel free to just post more specifically where you’re stuck here. If you have something working up to a point, that’s a huge help. If you don’t, I’d suggest simplifying the model until you do and then building up from there.
I have an implementation of Plackett-Luce in the following repo, which I used for estimating sushi preference given ranking (and rating) data:
It’s my StanCon submission for this Fall :-).
Is the stochastic process because the coefficients change over time?
1 Like
Thank you @ssp3nc3r and @Bob_Carpenter for your replies!
Someone reached out by email and was able to help solve my issue.