Dynamic estimation with random forests for discrete‐time survival data

Moradian, Hoora; Yao, Weichi; Larocque, Denis; Simonoff, Jeffrey S.; Frydman, Halina

doi:10.1002/cjs.11639

Cited by 9 publications

(5 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In future research the computation of standard errors for the parameters γ or for the hazards λ(t| x i ) directly, for example by bootstrap procedures, needs to be investigated. Moreover, the proposed class of models can be extended to an ensemble method, as survival forests for continuous (Ishwaran et al 2008;Moradian et al 2017;Wang et al 2018) and discrete-time data (Bou-Hamad et al 2011b;Schmid et al 2020;Moradian et al 2021), and adapted to competing risk data, similar to the approach by .…”

Section: Summary and Discussionmentioning

confidence: 99%

“…In addition, Tiendrébéogo et al (2019) applied a model-based recursive partitioning approach based on the algorithm by Hothorn et al (2006) to HIV patient data to identify characteristics that are associated with risk of death. For overviews on existing tree-structured methods for discrete-time hazard modeling and extensions for dynamic predictions, see also Kretowska (2019) and Moradian et al (2021).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Flexible tree-structured regression models for discrete event times

Spuck

Schmid²,

Heim³

et al. 2022

Stat Comput

View full text Add to dashboard Cite

Discrete hazard models are widely applied for the analysis of time-to-event outcomes that are intrinsically discrete or grouped versions of continuous event times. Commonly, one assumes that the effect of explanatory variables on the hazard can be described by a linear predictor function. This, however, may be not appropriate when non-linear effects or interactions between the explanatory variables occur in the data. To address this issue, we propose a novel class of discrete hazard models that utilizes recursive partitioning techniques and allows to include the effects of explanatory variables in a flexible data-driven way. We introduce a tree-building algorithm that inherently performs variable selection and facilitates the inclusion of non-linear effects and interactions, while the favorable additive form of the predictor function is kept. In a simulation study, the proposed class of models is shown to be competitive with alternative approaches, including a penalized parametric model and Bayesian additive regression trees, in terms of predictive performance and the ability to detect informative variables. The modeling approach is illustrated by two real-world applications analyzing data of patients with odontogenic infection and lymphatic filariasis.

show abstract

Section: Summary and Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Flexible tree-structured regression models for discrete event times

Spuck

Schmid²,

Heim³

et al. 2022

Stat Comput

View full text Add to dashboard Cite

show abstract

“…Other approaches (such as [35] and more recently [32]) designed longitudinal trees that use lagged response values as potential predictors, but still do not treat either the outcome or the covariates as inherently dynamic with time. Overall, in these methods, information is lost during the process, and the number of measurements per subject in real datasets can be too small to obtain satisfying regression parameters.…”

Section: Existing Longitudinal Tree-based Algorithmsmentioning

confidence: 99%

Time-penalized trees (TpT): a new tree-based data mining algorithm for time-varying covariates

Valla

2023

Preprint

View full text Add to dashboard Cite

This article proposes a new decision tree algorithm that accounts for time-varying covari-ates in the decision-making process. Traditional decision tree algorithms assume that the covariates are static and do not change over time, which can lead to inaccurate predictions in dynamic environments. Other existing methods suggest workaround solutions such as the pseudo-subject approach, discussed in the article. The proposed algorithm utilizes a different structure and a time-penalized splitting criterion that allow a recursive partitioning of both the covariates space and time. Relevant historical trends are then inherently involved in the construction of a tree, and are visible and interpretable once it is fit. This approach allows for innovative and highly interpretable analysis in settings where the covariates are subject to change over time. The effectiveness of the algorithm is demonstrated through real-world data analysis, highlighting its potential applications in various fields, including healthcare, finance, insurance, environmental monitoring, and data mining in general.

show abstract

“…10,11 There exist other survival trees and forest methods that can handle time-varying covariate data, but only for discrete-time survival data. [12][13][14][15][16] In this paper, we focus on forest algorithms for dynamic estimation of the survival function for continuous-time survival data. Ensemble methods like forest algorithms are known to preserve low bias while reducing variance and therefore can substantially improve prediction accuracy, compared to tree algorithms.…”

Section: Introductionmentioning

confidence: 99%

Ensemble methods for survival function estimation with time-varying covariates

Yao

Frydman

Larocque

et al. 2022

Stat Methods Med Res

Self Cite

View full text Add to dashboard Cite

Survival data with time-varying covariates are common in practice. If relevant, they can improve on the estimation of a survival function. However, the traditional survival forests—conditional inference forest, relative risk forest and random survival forest—have accommodated only time-invariant covariates. We generalize the conditional inference and relative risk forests to allow time-varying covariates. We also propose a general framework for estimation of a survival function in the presence of time-varying covariates. We compare their performance with that of the Cox model and transformation forest, adapted here to accommodate time-varying covariates, through a comprehensive simulation study in which the Kaplan-Meier estimate serves as a benchmark, and performance is compared using the integrated [Formula: see text] difference between the true and estimated survival functions. In general, the performance of the two proposed forests substantially improves over the Kaplan-Meier estimate. Taking into account all other factors, under the proportional hazard setting, the best method is always one of the two proposed forests, while under the non-proportional hazard setting, it is the adapted transformation forest. [Formula: see text]-fold cross-validation is used as an effective tool to choose between the methods in practice.

show abstract

Dynamic estimation with random forests for discrete‐time survival data

Cited by 9 publications

References 38 publications

Flexible tree-structured regression models for discrete event times

Flexible tree-structured regression models for discrete event times

Time-penalized trees (TpT): a new tree-based data mining algorithm for time-varying covariates

Ensemble methods for survival function estimation with time-varying covariates

Contact Info

Product

Resources

About