1
$\begingroup$

I have built a random survival forest using R package randomForestSRC. The OOB error rate is around 10%. I was wondering whether anyone had some experience in utilizing the outputs from this model for producing the "predicted future survival time"?

Thanks.

$\endgroup$

1 Answer 1

2
$\begingroup$

In the past, the authors of the randomForestSRC package have published a couple of articles on the subject that you may find useful for developing predicted survival curves.

For example, in this article, the authors provide some R code used to "predict" survival curves for a new set of data that they created on 3 made-up individuals to assess the "accuracy" of predicted survival curves.

On a side note, I have been unsuccessful in finding much literature on the use of random forests for survival curve prediction; hence, I am wary about applying such modeling to real-world applications - especially, since it is a paradigm shift from inferential to predictive application.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.