search

loft-br / xgboost-survival-embeddings

Improving XGBoost survival analysis with embeddings and debiased estimators
https://loft-br.github.io/xgboost-survival-embeddings/
Apache License 2.0
313 stars 51 forks source link

Competing risks/events #41

Closed onacrame closed 2 years ago

onacrame commented 2 years ago

If I have competing events and I treat the occurrence of a competing events as censored observations (informative/dependent censoring) will this model still give me an unbiased estimate of survival? In my use case the competing event will be one that precludes the event of interest, so the two events are completely mutually exclusive, I'm interested in event A, but if event B happens, then event A can never happen.

Generally I gather that competing events can work within a Cox framework, but it will give a biased estimate of incidence using a Kaplan Meier approach and XGBSE partially relies on Kaplan Meier.

GabrielGimenez commented 2 years ago

Given the informative censoring XGBSEKaplanNeighbors and XGBSEKaplanTree modules would return biased estimatives, but XGBSEDebiasedBCE should work just fine!

onacrame commented 2 years ago

Great thanks!

GabrielGimenez commented 2 years ago

@onacrame I'm not sure the survival curve will be meaningful tho, you might have to use return_interval_probs=True on predict and interpret your interval probabilities.

Actually, I don't know what would happen in your use case and since the XGBSEDebiasedBCE approximate a kaplan meier it might still return biased estimations

onacrame commented 2 years ago

Noted thanks.