2017
DOI: 10.1186/s12874-017-0383-8
|View full text |Cite
|
Sign up to set email alerts
|

A comparison of the conditional inference survival forest model to random survival forests based on a simulation study as well as on two applications with time-to-event data

Abstract: BackgroundRandom survival forest (RSF) models have been identified as alternative methods to the Cox proportional hazards model in analysing time-to-event data. These methods, however, have been criticised for the bias that results from favouring covariates with many split-points and hence conditional inference forests for time-to-event data have been suggested. Conditional inference forests (CIF) are known to correct the bias in RSF models by separating the procedure for the best covariate to split on from th… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
72
0

Year Published

2017
2017
2024
2024

Publication Types

Select...
7
1
1
1

Relationship

1
9

Authors

Journals

citations
Cited by 68 publications
(73 citation statements)
references
References 41 publications
1
72
0
Order By: Relevance
“…Given the fact that most of our covariates were categorical with more than two categorises, biased results on estimates such as variable importance are inevitable [ 53 , 55 ]. Our recent study[ 56 ] has therefore recommended the use of conditional inference forests suggested by [ 57 ] in the presence of covariates with many split points.…”
Section: Discussionmentioning
confidence: 99%
“…Given the fact that most of our covariates were categorical with more than two categorises, biased results on estimates such as variable importance are inevitable [ 53 , 55 ]. Our recent study[ 56 ] has therefore recommended the use of conditional inference forests suggested by [ 57 ] in the presence of covariates with many split points.…”
Section: Discussionmentioning
confidence: 99%
“…CF is an ensemble model with multiple conditional inference trees. We applied CF because there were more polytomous predictors than dichotomous predictors (the number of polytomous and dichotomous predictors was 9 and 3, respectively) in our dataset and CF has been shown as superior in predictive performance to RF on time-to-event datasets with polytomous predictors [ 34 ].…”
Section: Methodsmentioning
confidence: 99%
“…1 From 1 to number of trees (a) Take a bootstrap sample (b) Build tree on sample 2 Aggregate results from all trees The conditional inference forest (CIF) by [7] which builds each tree testing a global hypothesis is chosen over the random survival forest by [9] which adheres strictly to the random forest conditions laid out by [30], because it performs at least as well as the latter [37].…”
Section: Random Forest For Survival Datamentioning
confidence: 99%