Closed brunocarlin closed 1 year ago
Hi! :wave:
Thanks for the issue! I'd say it makes sense to put the port to the Python package in your own R package and then we can look into using that as an engine in censored. Regarding stacks: if the models are parsnip models, then stacking with {stacks} should work out of the box.
This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.
xgbse is a package from Python that uses base survival models cox and aft from xgboost and enhances them with 4 techniques
https://loft-br.github.io/xgboost-survival-embeddings/how_xgbse_works.html
XGBSEDebiasedBCE XGBSEStackedWeibull XGBSEKaplanNeighbors XGBSEKaplanTree
And allows you to bootstrap any of them using
XGBSEBootstrapEstimator
I want to develop a port into censored. I have already started the port of cox and aft
Should I split my work into my package, or can I keep it here?
Do you have any tips if I should focus on integrating with stacks?