pith. sign in

arxiv: 2406.04098 · v2 · submitted 2024-06-06 · 📊 stat.ML · cs.LG

A Large-Scale Neutral Comparison Study of Survival Models on Low-Dimensional Data

Lukas Burk , John Zobolas , Bernd Bischl , Andreas Bender , Marvin N. Wright , Raphael Sonabend This is my paper

Pith reviewed 2026-05-24 00:14 UTC · model grok-4.3

classification 📊 stat.ML cs.LG
keywords survival analysisbenchmarkCox proportional hazardsmachine learningright-censored datapredictive performancelow-dimensional datamodel comparison
0
0 comments X

The pith

No survival model significantly outperforms the Cox proportional hazards model on low-dimensional right-censored data.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper runs the first large-scale neutral benchmark of 19 survival models on 34 public low-dimensional datasets. Models are tuned separately with Harrell's C-index and the Integrated Survival Brier Score, then scored on six metrics that cover discrimination, calibration, and overall performance. Some learners post better average ranks, yet statistical tests find no significant edge over the Cox model for either tuning choice. This result matters for practitioners who routinely choose complex machine learning methods in standard survival settings. The full code, data, and results are released for direct inspection.

Core claim

In this neutral comparison on single-event right-censored low-dimensional data, no method significantly outperforms the Cox proportional hazards model for either tuning measure; therefore the Cox model remains a simple and robust choice sufficient for most practitioners.

What carries the argument

Large-scale benchmark that tunes and evaluates 19 models across 34 datasets using C-index and Integrated Survival Brier Score tuning plus six downstream metrics.

If this is right

  • Oblique random survival forests and likelihood-based boosting achieve the best average ranks for overall predictive performance.
  • Boosting, tree-based methods, and parametric models rank highest on discrimination.
  • The Cox model is sufficient for predictive tasks in the standard low-dimensional right-censored setting.
  • Results support using the Cox model as a default unless domain-specific evidence shows otherwise.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The conclusion may not extend to high-dimensional, competing-risks, or time-varying covariate settings.
  • Different calibration or time-dependent metrics could alter the ranking of methods.
  • The benchmark supplies a reusable reference set for testing future survival models under the same protocol.

Load-bearing premise

The 34 chosen datasets together with the selected tuning and evaluation protocol form a representative sample of typical low-dimensional right-censored survival problems.

What would settle it

A new collection of low-dimensional right-censored datasets or an alternative metric set on which at least one non-Cox model shows statistically significant improvement after identical tuning.

Figures

Figures reproduced from arXiv: 2406.04098 by Andreas Bender, Bernd Bischl, John Zobolas, Lukas Burk, Marvin N. Wright, Raphael Sonabend.

Figure 1
Figure 1. Figure 1: Critical difference plot comparing models with the CPH reference tuned on Harrell’s C (a,b) [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Boxplots of aggregated scores across all datasets for models tuned and evaluated with [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Boxplots of raw evaluation scores using discrimination measures for tuning (Harrell’s C) [PITH_FULL_IMAGE:figures/full_fig_p028_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Boxplots of scaled evaluation scores using discrimination measures for tuning (Harrell’s C) [PITH_FULL_IMAGE:figures/full_fig_p029_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Boxplots of raw evaluation scores using scoring rules for tuning (RCLL) and evaluation. [PITH_FULL_IMAGE:figures/full_fig_p030_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Boxplots of ERV evaluation scores using scoring rules for tuning (RCLL) and evaluation. [PITH_FULL_IMAGE:figures/full_fig_p031_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Raw scores for Harrell’s C across outer resampling folds per dataset, task, and evaluation [PITH_FULL_IMAGE:figures/full_fig_p033_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Raw scores for Uno’s C across outer resampling folds per dataset, task, and evaluation [PITH_FULL_IMAGE:figures/full_fig_p034_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Raw scores for ISBS across outer resampling folds per dataset, task, and evaluation measure [PITH_FULL_IMAGE:figures/full_fig_p035_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Raw scores for RCLL across outer resampling folds per dataset, task, and evaluation [PITH_FULL_IMAGE:figures/full_fig_p036_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Raw scores for ISBS across outer resampling folds per dataset, task, and evaluation [PITH_FULL_IMAGE:figures/full_fig_p037_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Raw scores for RNLL across outer resampling folds per dataset, task, and evaluation [PITH_FULL_IMAGE:figures/full_fig_p038_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Raw scores for RISLL across outer resampling folds per dataset, task, and evaluation [PITH_FULL_IMAGE:figures/full_fig_p039_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: D-Calibration p-value heatmaps across all datasets and learners. ‘X’ indicates [PITH_FULL_IMAGE:figures/full_fig_p040_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Calibration scores using van Houwelingen’s [PITH_FULL_IMAGE:figures/full_fig_p041_15.png] view at source ↗
read the original abstract

This work presents the first large-scale neutral benchmark experiment focused on single-event, right-censored, low-dimensional survival data. Benchmark experiments are essential in methodological research to scientifically compare new and existing model classes through proper empirical evaluation. Existing benchmarks in the survival literature are smaller in scale regarding the number of used datasets and extent of empirical evaluation. They often lack appropriate tuning or evaluation procedures, while other comparison studies focus on qualitative reviews rather than quantitative comparisons. This comprehensive study aims to fill the gap by neutrally evaluating a broad range of methods and providing generalizable guidelines for practitioners. We benchmark 19 models, ranging from classical statistical approaches to many common machine learning methods, on 34 publicly available datasets. The benchmark tunes models using both a discrimination measure (Harrell's C-index) and a scoring rule (Integrated Survival Brier Score), and evaluates them across six metrics covering discrimination, calibration, and overall predictive performance. Despite superior average ranks in overall predictive performance from individual learners like oblique random survival forests and likelihood-based boosting, and better discrimination rankings from multiple boosting- and tree-based methods as well as parametric survival models, no method significantly outperforms the commonly used Cox proportional hazards model for either tuning measure. We conclude that for predictive purposes in the standard survival analysis setting of low-dimensional, right-censored data, the Cox Proportional Hazards model remains a simple and robust method, sufficient for most practitioners. All code, data, and results are publicly available on GitHub https://github.com/slds-lmu/paper_2023_survival_benchmark

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 1 minor

Summary. The paper presents the first large-scale neutral benchmark of 19 survival models (classical statistical to ML) on 34 publicly available low-dimensional, right-censored, single-event datasets. Models are tuned separately using Harrell's C-index and Integrated Survival Brier Score, then evaluated on six metrics spanning discrimination, calibration, and overall performance. The central empirical result is that, despite some methods showing better average ranks, no method significantly outperforms the Cox proportional hazards model under either tuning protocol. The authors conclude that Cox PH remains a simple, robust, and sufficient choice for most practitioners in this setting. All code, data, and results are released publicly on GitHub.

Significance. If the central null result generalizes, the study supplies a valuable, large-scale empirical anchor for the survival-analysis literature: it indicates that, in the standard low-dimensional right-censored regime, gains from more complex learners are not statistically detectable under proper tuning and multiple metrics. The explicit public release of code, data, and full results is a clear strength that supports reproducibility and secondary analyses.

major comments (3)
  1. [Abstract / dataset section] Abstract and dataset description: the claim that results support the broad conclusion that 'Cox PH remains ... sufficient for most practitioners' is load-bearing on the assumption that the 34 datasets constitute a representative sample of low-dimensional right-censored problems. The manuscript provides no sampling frame, inclusion/exclusion criteria, or analysis of selection effects (e.g., publication bias toward datasets where linear PH models already perform well), nor does it report the empirical distribution of censoring rates, event frequencies, or covariate correlations across the collection.
  2. [Results / statistical analysis] Results section (statistical testing): the statement that 'no method significantly outperforms' the Cox model requires explicit description of the paired statistical test, the exact significance threshold, and any correction for multiple comparisons across 19 models and 6 metrics. Without these details the reported lack of significance cannot be verified and directly affects the central claim.
  3. [Methods / evaluation protocol] Evaluation protocol: the paper tunes on C-index or ISBS and evaluates on six metrics, yet does not report sensitivity of the ranking conclusions to the choice of tuning measure or to the handling of ties and censoring in the C-index. Because the null result is asserted for 'either tuning measure,' this omission is material to the robustness of the headline finding.
minor comments (1)
  1. [Abstract] The GitHub repository link is given but the manuscript does not state the exact commit or tag corresponding to the results reported in the paper.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive comments, which help clarify the scope and robustness of our benchmark. We address each major point below and indicate where revisions will be made.

read point-by-point responses

  1. Referee: [Abstract / dataset section] Abstract and dataset description: the claim that results support the broad conclusion that 'Cox PH remains ... sufficient for most practitioners' is load-bearing on the assumption that the 34 datasets constitute a representative sample of low-dimensional right-censored problems. The manuscript provides no sampling frame, inclusion/exclusion criteria, or analysis of selection effects (e.g., publication bias toward datasets where linear PH models already perform well), nor does it report the empirical distribution of censoring rates, event frequencies, or covariate correlations across the collection.

    Authors: We agree that a more explicit description of dataset selection strengthens the interpretation. The 34 datasets comprise all publicly available low-dimensional (p typically < 50), right-censored, single-event survival datasets meeting minimum size and completeness criteria that we could locate from standard repositories (UCI, OpenML, R packages, and published studies). We will add a dedicated subsection detailing the exact inclusion/exclusion criteria, a table or figure summarizing the empirical distributions of censoring rates, event frequencies, and pairwise covariate correlations, and a brief discussion of potential selection effects. This addition will better qualify the scope of our conclusions without altering the central empirical findings. revision: yes

  2. Referee: [Results / statistical analysis] Results section (statistical testing): the statement that 'no method significantly outperforms' the Cox model requires explicit description of the paired statistical test, the exact significance threshold, and any correction for multiple comparisons across 19 models and 6 metrics. Without these details the reported lack of significance cannot be verified and directly affects the central claim.

    Authors: The methods section already specifies the use of paired Wilcoxon signed-rank tests on per-dataset metric differences with a nominal threshold of 0.05; no multiplicity correction was applied because the primary interest is the direct comparison against Cox rather than an exhaustive ranking. However, we acknowledge that these details are not restated in the results narrative. We will insert a concise paragraph in the results section that repeats the test procedure, threshold, and rationale for the correction choice, along with a note that all pairwise p-values versus Cox remain above 0.05 under both tuning regimes. revision: yes

  3. Referee: [Methods / evaluation protocol] Evaluation protocol: the paper tunes on C-index or ISBS and evaluates on six metrics, yet does not report sensitivity of the ranking conclusions to the choice of tuning measure or to the handling of ties and censoring in the C-index. Because the null result is asserted for 'either tuning measure,' this omission is material to the robustness of the headline finding.

    Authors: The headline null result is already shown to hold under both tuning protocols separately. We will add a short sensitivity paragraph (or supplementary table) confirming that the ranking order versus Cox is stable when the C-index is replaced by its tie-adjusted variant and when censoring is handled via the standard inverse-probability weighting in the Brier score. A exhaustive grid of every possible tie/censoring variant is beyond the scope of the current study, but the two primary tuning measures already bracket the main practical choices. revision: partial

Circularity Check

0 steps flagged

No circularity: purely empirical benchmark with no derivations or self-referential fits

full rationale

This is an empirical benchmark study that directly evaluates 19 survival models on 34 public datasets using standard tuning (C-index, ISBS) and six evaluation metrics. No derivation chain, first-principles result, or prediction exists that could reduce to the paper's own inputs by construction. No self-definitional steps, fitted inputs renamed as predictions, or load-bearing self-citations are present; conclusions follow from observed performance ranks and significance tests on external data. The representativeness concern raised by the skeptic is a generalizability issue, not circularity under the enumerated patterns.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on the representativeness of the chosen datasets and the fairness of the benchmarking protocol rather than new mathematical constructs or fitted parameters.

axioms (2)
  • domain assumption The 34 publicly available datasets are representative of typical low-dimensional, single-event, right-censored survival problems.
    Generalization from the benchmark to 'most practitioners' depends on this.
  • domain assumption The dual tuning (C-index and ISBS) plus six-metric evaluation protocol provides an unbiased comparison across model classes.
    The claim of 'neutral' comparison and the null result on significance rest on this protocol being appropriate.

pith-pipeline@v0.9.0 · 5825 in / 1537 out tokens · 30394 ms · 2026-05-24T00:14:00.071366+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

127 extracted references · 127 canonical work pages · 2 internal anchors

  1. [1]

    Nonparametric Inference for a Family of Counting Processes

    Odd Aalen. “Nonparametric Inference for a Family of Counting Processes”. In: The Annals of Statistics 6.4 (1978), pp. 701–726

    work page 1978

  2. [2]

    Nearest Neighbor Estimation of a Bivariate Distribution Under Random Censoring

    Michael G Akritas. “Nearest Neighbor Estimation of a Bivariate Distribution Under Random Censoring”. en. In: Ann. Statist. 22.3 (1994), pp. 1299–1327. ISSN : 0090-5364. DOI: 10. 1214 / aos / 1176325630. URL: https : / / projecteuclid . org : 443 / euclid . aos / 1176325630

    work page 1994

  3. [3]

    Nonalcoholic fatty liver disease incidence and impact on metabolic burden and death: A 20 year-community study

    Alina M Allen et al. “Nonalcoholic fatty liver disease incidence and impact on metabolic burden and death: A 20 year-community study.” eng. In:Hepatology (Baltimore, Md.) 67.5 (Apr. 2018), pp. 1726–1736. ISSN : 1527-3350 (Electronic). DOI: 10.1002/hep.29546

    work page doi:10.1002/hep.29546 2018

  4. [4]

    Introduction

    Per Kragh Andersen et al. “Introduction”. In:Statistical Models Based on Counting Processes. Springer Series in Statistics. New York, NY: Springer US, 1993, pp. 1–44. DOI: https: //doi.org/10.1007/978-1-4612-4348-9

    work page doi:10.1007/978-1-4612-4348-9 1993

  5. [5]

    Austin, Frank E

    Peter C. Austin, Frank E. Harrell Jr, and David van Klaveren. “Graphical Calibration Curves and the Integrated Calibration Index (ICI) for Survival Models”. In:Statistics in Medicine 39.21 (2020), pp. 2714–2742. ISSN : 1097-0258. DOI: 10.1002/sim.8570

    work page doi:10.1002/sim.8570 2020

  6. [6]

    Countdown Regression: Sharp and Calibrated Survival Predictions

    Anand Avati et al. “Countdown Regression: Sharp and Calibrated Survival Predictions”. In: Proceedings of The 35th Uncertainty in Artificial Intelligence Conference. Uncertainty in Artificial Intelligence. PMLR, Aug. 6, 2020, pp. 145–155. URL: https://proceedings. mlr.press/v115/avati20a.html (visited on 05/21/2024)

    work page 2020

  7. [7]

    Survival Regression with Accelerated Failure Time Model in XGBoost

    Avinash Barnwal, Hyunsu Cho, and Toby Hocking. “Survival Regression with Accelerated Failure Time Model in XGBoost”. In:Journal of Computational and Graphical Statistics31.4 (Oct. 2, 2022), pp. 1292–1302. ISSN : 1061-8600. DOI: 10.1080/10618600.2022.2067548

    work page doi:10.1080/10618600.2022.2067548 2022

  8. [8]

    Advanced Research Computing Center

    Beartooth Computing Environment, x86_64 cluster. Advanced Research Computing Center. University of Wyoming, Laramie, WY. 2024. DOI: {https : / / doi . org / 10 . 15786 / M2FY47}

    work page 2024

  9. [9]

    mlr3tuningspaces: Search Spaces for ’mlr3’

    Marc Becker. mlr3tuningspaces: Search Spaces for ’mlr3’. GitHub, 2024. URL: https : //github.com/mlr-org/mlr3extralearners

    work page 2024

  10. [10]

    pammtools: Piece-wise exponential Additive Mixed Modeling tools

    Andreas Bender and Fabian Scheipl. “pammtools: Piece-wise exponential Additive Mixed Modeling tools”. In: arXiv:1806.01042 [stat] (2018). URL: http://arxiv.org/abs/1806. 01042

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  11. [11]

    Penalized estimation of complex, non-linear exposure-lag-response associations

    Andreas Bender et al. “Penalized estimation of complex, non-linear exposure-lag-response associations”. In: Biostatistics 20.2 (Feb. 2018), pp. 315–331. ISSN : 1465-4644. DOI: 10. 1093/biostatistics/kxy003 . URL: https://doi.org/10.1093/biostatistics/ kxy003

    work page doi:10.1093/biostatistics/ 2018

  12. [12]

    Allowing for Mandatory Covariates in Boosting Estimation of Sparse High-Dimensional Survival Models

    Harald Binder and Martin Schumacher. “Allowing for Mandatory Covariates in Boosting Estimation of Sparse High-Dimensional Survival Models”. In: BMC Bioinformatics 9.1 (Jan. 10, 2008), p. 14. ISSN : 1471-2105. DOI: 10.1186/1471-2105-9-14 . 10

    work page doi:10.1186/1471-2105-9-14 2008

  13. [13]

    mlr3pipelines - Flexible Machine Learning Pipelines in R

    Martin Binder et al. “mlr3pipelines - Flexible Machine Learning Pipelines in R”. In: Journal of Machine Learning Research 22.184 (2021), pp. 1–7. URL: http://jmlr.org/papers/ v22/21-0281.html

    work page 2021

  14. [14]

    Hyperparameter Optimization: Foundations, Algorithms, Best Practices, and Open Challenges

    Bernd Bischl et al. “Hyperparameter Optimization: Foundations, Algorithms, Best Practices, and Open Challenges”. In:WIREs Data Mining and Knowledge Discovery13.2 (2023), e1484. ISSN : 1942-4795. DOI: 10.1002/widm.1484

    work page doi:10.1002/widm.1484 2023

  15. [15]

    Estimating and Com- paring Time-Dependent Areas under Receiver Operating Characteristic Curves for Censored Event Times with Competing Risks

    Paul Blanche, Jean-François Dartigues, and Hélène Jacqmin-Gadda. “Estimating and Com- paring Time-Dependent Areas under Receiver Operating Characteristic Curves for Censored Event Times with Competing Risks”. In: Statistics in Medicine 32.30 (2013), pp. 5381–5397. ISSN : 1097-0258. DOI: 10.1002/sim.5958

    work page doi:10.1002/sim.5958 2013

  16. [16]

    A Plea for Neutral Com- parison Studies in Computational Sciences

    Anne-Laure Boulesteix, Sabine Lauer, and Manuel J. A. Eugster. “A Plea for Neutral Com- parison Studies in Computational Sciences”. In: PLOS ONE 8.4 (Apr. 24, 2013), e61562. ISSN : 1932-6203. DOI: 10.1371/journal.pone.0061562

    work page doi:10.1371/journal.pone.0061562 2013

  17. [17]

    Classification and Regression Trees

    Leo Breiman, ed. Classification and Regression Trees. 1. CRC Press repr. Boca Raton, Fla.: Chapman & Hall/CRC, 1998. 358 pp. ISBN : 978-0-412-04841-8

    work page 1998

  18. [18]

    Design and Analysis of Two-Phase Studies with Binary Outcome Applied to Wilms Tumour Prognosis

    N. E. Breslow and N. Chatterjee. “Design and Analysis of Two-Phase Studies with Binary Outcome Applied to Wilms Tumour Prognosis”. In:Journal of the Royal Statistical Society Series C: Applied Statistics 48.4 (Dec. 1, 1999), pp. 457–468. ISSN : 0035-9254. DOI: 10. 1111/1467-9876.00165

    work page arXiv 1999

  19. [19]

    eha: Event History Analysis

    Göran Broström. eha: Event History Analysis. R package version 2.9.0. 2021. URL: http: //ehar.se/r/eha/

    work page 2021

  20. [20]

    Boosting With the L2 Loss: Regression and Classification

    Peter Bühlmann and Bin Yu. “Boosting With the L2 Loss: Regression and Classification”. In: Journal of the American Statistical Association 98.462 (June 1, 2003), pp. 324–339. ISSN : 0162-1459. DOI: 10.1198/016214503000125

    work page doi:10.1198/016214503000125 2003

  21. [21]

    smcure: Fit Semiparametric Mixture Cure Models

    Chao Cai et al. smcure: Fit Semiparametric Mixture Cure Models. R package version 2.0

    work page

  22. [22]

    URL: https://CRAN.R-project.org/package=smcure

    work page

  23. [23]

    Supplemental Materials to Bayesian Methods for Data Analysis, 3rd Edition

    Bradley P. Carlin and Thomas A. Louis. “Supplemental Materials to Bayesian Methods for Data Analysis, 3rd Edition”. In: (Oct. 2, 2018). DOI: 10.13020/D6N10N

    work page doi:10.13020/d6n10n 2018

  24. [24]

    Groups, the Media, Agency Waiting Costs, and FDA Drug Approval

    Daniel P. Carpenter. “Groups, the Media, Agency Waiting Costs, and FDA Drug Approval”. In: American Journal of Political Science 46.3 (2002), pp. 490–505. ISSN : 00925853, 15405907. URL: http://www.jstor.org/stable/3088394

    work page arXiv 2002

  25. [25]

    An Empirical Comparison of Supervised Learning Algorithms

    Rich Caruana and Alexandru Niculescu-Mizil. “An Empirical Comparison of Supervised Learning Algorithms”. In: Proceedings of the 23rd International Conference on Machine Learning - ICML ’06. The 23rd International Conference. Pittsburgh, Pennsylvania: ACM Press, 2006, pp. 161–168. ISBN : 978-1-59593-383-6. DOI: 10.1145/1143844.1143865

    work page doi:10.1145/1143844.1143865 2006

  26. [26]

    Chen and C

    Tianqi Chen and Carlos Guestrin. “XGBoost: A Scalable Tree Boosting System”. In:Proceed- ings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. KDD ’16. New York, NY , USA: Association for Computing Machinery, Aug. 13, 2016, pp. 785–794. ISBN : 978-1-4503-4232-2. DOI: 10.1145/2939672.2939785

    work page doi:10.1145/2939672.2939785 2016

  27. [27]

    Partial Likelihood

    D. R. Cox. “Partial Likelihood”. In: Biometrika 62.2 (Aug. 1, 1975), pp. 269–276. ISSN : 0006-3444. DOI: 10.1093/biomet/62.2.269

    work page doi:10.1093/biomet/62.2.269 1975

  28. [28]

    Statistical comparisons of classifiers over multiple data sets

    Janez Demšar. “Statistical comparisons of classifiers over multiple data sets”. In:Journal of Machine learning research 7.1 (2006), pp. 1–30

    work page 2006

  29. [29]

    Time to default in credit scoring using survival analysis: A benchmark study

    Lore Dirick, Gerda Claeskens, and Bart Baesens. “Time to default in credit scoring using survival analysis: A benchmark study”. In: Journal of the Operational Research Society 68.6 (2017), pp. 652–665. ISSN : 14769360. DOI: 10.1057/s41274-016-0128-9

    work page doi:10.1057/s41274-016-0128-9 2017

  30. [30]

    doi: 10.1016/j.mayocp.2012.03.009

    Angela Dispenzieri et al. “Use of nonclonal serum immunoglobulin free light chains to predict overall survival in the general population”. eng. In: Mayo Clinic proceedings 87.6 (June 2012), pp. 517–523. ISSN : 1942-5546. DOI: 10.1016/j.mayocp.2012.03.009. URL: https://www.ncbi.nlm.nih.gov/pubmed/22677072%20https://www.ncbi.nlm. nih.gov/pmc/articles/PMC3538473/

    work page doi:10.1016/j.mayocp.2012.03.009 2012

  31. [31]

    SurvSet: An Open-Source Time-to-Event Dataset Repository

    Erik Drysdale. “SurvSet: An Open-Source Time-to-Event Dataset Repository”. Mar. 6, 2022. arXiv: 2203.03094 [cs, stat] . URL: http://arxiv.org/abs/2203.03094 (visited on 03/12/2022)

    work page arXiv 2022

  32. [32]

    Large-Scale Benchmarking

    Sebastian Fischer, Michel Lang, and Marc Becker. “Large-Scale Benchmarking”. In:Applied Machine Learning Using mlr3 in R . Ed. by Bernd Bischl et al. CRC Press, 2024. URL: https://mlr3book.mlr-org.com/large-scale_benchmarking.html. 11

    work page 2024

  33. [33]

    Foucher et al

    Y . Foucher et al. RISCA: Causal Inference and Prediction in Cohort-Based Analyses. R package version 1.0.4. 2023. URL: https://CRAN.R-project.org/package=RISCA

    work page 2023

  34. [34]

    Stochastic Gradient Boosting

    Jerome Friedman. “Stochastic Gradient Boosting”. In: Computational Statistics & Data Analysis 38 (Mar. 1999), pp. 367–378. DOI: 10.1016/S0167-9473(01)00065-2

    work page doi:10.1016/s0167-9473(01)00065-2 1999

  35. [35]

    Bayesian Optimization

    Roman Garnett. Bayesian Optimization. Cambridge University Press, 2023

    work page 2023

  36. [36]

    Comparisons between Survival Models in Predicting Cardiovascular Disease Events : Application in the ATTICA Study ( 2002-2012 )

    Ekavi N Georgousopoulou et al. “Comparisons between Survival Models in Predicting Cardiovascular Disease Events : Application in the ATTICA Study ( 2002-2012 ).” In: Journal of Statistics Applications & Probability 4.2 (2015), pp. 203–210

    work page 2002

  37. [37]

    L1 Penalized Estimation in the Cox Proportional Hazards Model

    Jelle J. Goeman. “L1 Penalized Estimation in the Cox Proportional Hazards Model”. In: Biometrical Journal 52.1 (2010), pp. 70–84. ISSN : 1521-4036. DOI: 10 . 1002 / bimj . 200900028

    work page 2010

  38. [38]

    Performance Evaluation of Support Vector Regression Models for Survival Analysis: A Simulation Study

    Shahrbanoo Goli et al. “Performance Evaluation of Support Vector Regression Models for Survival Analysis: A Simulation Study”. In:International Journal of Advanced Computer Science and Applications 7.6 (2016). ISSN : 21565570, 2158107X. DOI: 10.14569/IJACSA. 2016.070650

    work page doi:10.14569/ijacsa 2016

  39. [39]

    Assessment and Comparison of Prognostic Classification Schemes for Survival Data

    E. Graf et al. “Assessment and Comparison of Prognostic Classification Schemes for Survival Data”. In: Statistics in Medicine 18.17-18 (Sept. 15, 1999), pp. 2529–2545. ISSN : 0277-

    work page 1999

  40. [40]

    1002 / (sici ) 1097 - 0258(19990915 / 30 ) 18 : 17 / 18<2529 :: aid - sim274>3.0.co;2-5

    DOI: 10 . 1002 / (sici ) 1097 - 0258(19990915 / 30 ) 18 : 17 / 18<2529 :: aid - sim274>3.0.co;2-5. pmid: 10474158

    work page

  41. [41]

    frailtyHL: Frailty Models via Hierarchical Likelihood

    Il Do Ha et al. frailtyHL: Frailty Models via Hierarchical Likelihood. R package version 2.3

    work page

  42. [42]

    URL: https://CRAN.R-project.org/package=frailtyHL

    work page

  43. [43]

    Comparison of Survival Models for Analyzing Prognostic Factors in Gastric Cancer Patients

    Danial Habibi et al. “Comparison of Survival Models for Analyzing Prognostic Factors in Gastric Cancer Patients”. In: Asian Pacific journal of cancer prevention : APJCP 19.3 (2018), pp. 749–753. ISSN : 2476-762X. DOI: 10.22034/APJCP.2018.19.3.749 . URL: http://www.ncbi.nlm.nih.gov/pubmed/29582630%7B%5C%%7D0Ahttp://www. pubmedcentral.nih.gov/articlerender....

    work page doi:10.22034/apjcp.2018.19.3.749 2018

  44. [44]

    Effective Ways to Build and Evaluate Individual Survival Distributions

    Humza Haider et al. “Effective Ways to Build and Evaluate Individual Survival Distributions”. In: Journal of Machine Learning Research 21.85 (2020), pp. 1–63. ISSN : 1533-7928. URL: http://jmlr.org/papers/v21/18-772.html (visited on 05/21/2024)

    work page 2020

  45. [45]

    Evaluating the yield of medical tests

    Frank E. Harrell, Robert M. Califf, and David B. Pryor. “Evaluating the yield of medical tests”. In: JAMA 247.18 (May 1982), pp. 2543–2546. ISSN : 0098-7484. URL: http://dx. doi.org/10.1001/jama.1982.03320430047030

    work page doi:10.1001/jama.1982.03320430047030 1982

  46. [46]

    Large-Scale Benchmark Study of Survival Prediction Methods Using Multi-Omics Data

    Moritz Herrmann et al. “Large-Scale Benchmark Study of Survival Prediction Methods Using Multi-Omics Data”. In:Briefings in Bioinformatics 22.3 (May 1, 2021), bbaa167. ISSN : 1477-4054. DOI: 10.1093/bib/bbaa167

    work page doi:10.1093/bib/bbaa167 2021

  47. [47]

    Applied survival analysis regres- sion modeling of time-to-event data

    David W Hosmer, Stanley Lemeshow, and Susanne May. Applied survival analysis regres- sion modeling of time-to-event data. eng. 2nd ed. Wiley series in probability and statistics. Hoboken, N.J.: Wiley-Interscience, 2008

    work page 2008

  48. [48]

    Applied survival analysis: regression modeling of time-to-event data

    David W Hosmer Jr, Stanley Lemeshow, and Susanne May. Applied survival analysis: regression modeling of time-to-event data. V ol. 618. John Wiley & Sons, 2011

    work page 2011

  49. [49]

    Unbiased Recursive Partitioning: A Conditional Inference Framework

    Torsten Hothorn, Kurt Hornik, and Achim Zeileis. “Unbiased Recursive Partitioning: A Conditional Inference Framework”. In: Journal of Computational and Graphical Statistics 15.3 (Sept. 1, 2006), pp. 651–674. ISSN : 1061-8600. DOI: 10.1198/106186006X133933

    work page doi:10.1198/106186006x133933 2006

  50. [50]

    Validation, Calibration, Revision and Combination of Prognostic Survival Models

    Hans C. van Houwelingen. “Validation, Calibration, Revision and Combination of Prognostic Survival Models”. In: Statistics in Medicine 19.24 (2000), pp. 3401–3415. ISSN : 1097-0258. DOI: 10.1002/1097-0258(20001230)19:24<3401::AID-SIM554>3.0.CO;2-2

    work page doi:10.1002/1097-0258(20001230)19:24 2000

  51. [51]

    Random survival forests

    By Hemant Ishwaran et al. “Random survival forests”. In:The Annals of Statistics 2.3 (2008), pp. 841–860. DOI: 10.1214/08-AOAS169. arXiv: arXiv:0811.1645v1

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1214/08-aoas169 2008

  52. [52]

    Accelerated and Interpretable Oblique Random Survival Forests

    Byron C. Jaeger et al. “Accelerated and Interpretable Oblique Random Survival Forests”. In: Journal of Computational and Graphical Statistics 33.1 (Jan. 2, 2024), pp. 192–207. ISSN : 1061-8600. DOI: 10.1080/10618600.2023.2231048

    work page doi:10.1080/10618600.2023.2231048 2024

  53. [53]

    Acute Stroke With Atrial Fibrillation

    Henrik Stig Jørgensen et al. “Acute Stroke With Atrial Fibrillation”. In:Stroke 27.10 (Oct. 1, 1996), pp. 1765–1769. DOI: 10.1161/01.STR.27.10.1765

    work page doi:10.1161/01.str.27.10.1765 1996

  54. [54]

    Kalbfleisch and Ross L

    John D. Kalbfleisch and Ross L. Prentice. The Statistical Analysis of Failure Time Data. John Wiley & Sons, Jan. 25, 2011. 464 pp. ISBN : 978-1-118-03123-0. 12

    work page 2011

  55. [55]

    Nonparametric Estimation from Incomplete Observations

    E. L. Kaplan and Paul Meier. “Nonparametric Estimation from Incomplete Observations”. In: Journal of the American Statistical Association 53.282 (1958), pp. 457–481. ISSN : 01621459. DOI: 10.2307/2281868

    work page doi:10.2307/2281868 1958

  56. [56]

    Comparison of Cox Regression With Other Methods for Determining Prediction Models and Nomograms

    Michael W. Kattan. “Comparison of Cox Regression With Other Methods for Determining Prediction Models and Nomograms”. In: The Journal of Urology. Part 2 of 2 170 (6, Supple- ment Dec. 1, 2003), S6–S10. ISSN : 0022-5347. DOI: 10.1097/01.ju.0000094764.56269. 2d

    work page doi:10.1097/01.ju.0000094764.56269 2003

  57. [57]

    The Index of Prediction Accuracy: An Intuitive Measure Useful for Evaluating Risk Prediction Models

    Michael W. Kattan and Thomas A. Gerds. “The Index of Prediction Accuracy: An Intuitive Measure Useful for Evaluating Risk Prediction Models”. In: Diagnostic and Prognostic Research 2.1 (May 4, 2018), p. 7. ISSN : 2397-7523. DOI: 10.1186/s41512-018-0029-2

    work page doi:10.1186/s41512-018-0029-2 2018

  58. [58]

    DeepSurv

    Jared Katzman. DeepSurv. 2016. URL: https://pypi.org/project/deepsurv/

    work page 2016

  59. [59]

    Katzman, Uri Shaham, Alexander Cloninger, Jonathan Bates, Tingting Jiang, and Yuval Kluger

    Jared L Katzman et al. “DeepSurv: personalized treatment recommender system using a Cox proportional hazards deep neural network”. In: BMC Medical Research Methodology 18.1 (2018), p. 24. ISSN : 1471-2288. DOI: 10.1186/s12874-018-0482-1 . URL: https: //doi.org/10.1186/s12874-018-0482-1

    work page doi:10.1186/s12874-018-0482-1 2018

  60. [60]

    Interferon alfa-2b adjuvant therapy of high-risk resected cutaneous melanoma: the Eastern Cooperative Oncology Group Trial EST 1684

    John M Kirkwood et al. “Interferon alfa-2b adjuvant therapy of high-risk resected cutaneous melanoma: the Eastern Cooperative Oncology Group Trial EST 1684.” In:Journal of clinical oncology 14.1 (1996), pp. 7–17

    work page 1996

  61. [61]

    Survival analysis: techniques for censored and truncated data

    John P Klein and Melvin L Moeschberger. Survival analysis: techniques for censored and truncated data. 2nd ed. Springer Science & Business Media, 2003. ISBN : 0387216456

    work page 2003

  62. [62]

    quantreg: Quantile Regression

    Roger Koenker. quantreg: Quantile Regression. R package version 5.86. 2021. URL: https: //www.r-project.org

    work page 2021

  63. [63]

    Explained Residual Variation, Explained Risk, and Goodness of Fit

    Edward L Korn and Richard Simon. “Explained Residual Variation, Explained Risk, and Goodness of Fit”. In: The American Statistician 45.3 (Dec. 1991), pp. 201–206. ISSN : 00031305. DOI: 10.2307/2684290. URL: http://www.jstor.org/stable/2684290

    work page doi:10.2307/2684290 1991

  64. [64]

    Håvard Kvamme. pycox. 2018. URL: https://pypi.org/project/pycox/

    work page 2018

  65. [65]

    "Benign" monoclonal gammopathy–after 20 to 35 years of follow-up

    R A Kyle. “"Benign" monoclonal gammopathy–after 20 to 35 years of follow-up.” eng. In: Mayo Clinic proceedings 68.1 (Jan. 1993), pp. 26–36. ISSN : 0025-6196 (Print). DOI: 10.1016/s0025-6196(12)60015-9

    work page doi:10.1016/s0025-6196(12)60015-9 1993

  66. [66]

    Good Parameters and Implementations for Combined Multiple Recursive Random Number Generators

    Pierre L’Ecuyer. “Good Parameters and Implementations for Combined Multiple Recursive Random Number Generators”. In: Operations Research 47.1 (Feb. 1999), pp. 159–164. ISSN : 0030-364X. DOI: 10.1287/opre.47.1.159 . URL: https://pubsonline.informs. org/doi/abs/10.1287/opre.47.1.159

    work page doi:10.1287/opre.47.1.159 1999

  67. [67]

    Batchtools: Tools for R to Work on Batch Systems

    Michel Lang, Bernd Bischl, and Dirk Surmann. “Batchtools: Tools for R to Work on Batch Systems”. In: Journal of Open Source Software2.10 (Feb. 22, 2017), p. 135.ISSN : 2475-9066. DOI: 10.21105/joss.00135

    work page doi:10.21105/joss.00135 2017

  68. [68]

    mlr3: A modern object-oriented machine learning framework in R

    Michel Lang et al. “mlr3: A modern object-oriented machine learning framework in R”. In: Journal of Open Source Software 4.44 (2019), p. 1903. DOI: 10.21105/joss.01903. URL: https://joss.theoj.org/papers/10.21105/joss.01903

    work page doi:10.21105/joss.01903 2019

  69. [69]

    mlr3tuning: Tuning for ’mlr3’

    Michel Lang et al. mlr3tuning: Tuning for ’mlr3’. 2023. URL: https://cran.r-project. org/package=mlr3tuning

    work page 2023

  70. [70]

    Review of statistical methods for survival analysis using genomic data

    Seungyeoun Lee and Heeju Lim. “Review of statistical methods for survival analysis using genomic data”. eng. In: Genomics & informatics 17.4 (Dec. 2019), e41–e41. ISSN : 1598- 866X. DOI: 10.5808/GI.2019.17.4.e41 . URL: https://pubmed.ncbi.nlm.nih. gov/31896241%20https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6944043/

    work page doi:10.5808/gi.2019.17.4.e41 2019

  71. [71]

    Lee, Changhee et al. DeepHit. 2019. URL: https://github.com/chl8856/DeepHit

    work page 2019

  72. [72]

    On the Breslow estimator

    D. Y . Lin. “On the Breslow estimator”. In:Lifetime Data Analysis 13.4 (2007), pp. 471–480. ISSN : 13807870. DOI: 10.1007/s10985-007-9048-y

    work page doi:10.1007/s10985-007-9048-y 2007

  73. [73]

    Prospective evaluation of prognostic variables from patient-completed questionnaires. North Central Cancer Treatment Group

    C L Loprinzi et al. “Prospective evaluation of prognostic variables from patient-completed questionnaires. North Central Cancer Treatment Group.” eng. In:Journal of clinical oncology : official journal of the American Society of Clinical Oncology12.3 (Mar. 1994), pp. 601–607. ISSN : 0732-183X (Print). DOI: 10.1200/JCO.1994.12.3.601

    work page doi:10.1200/jco.1994.12.3.601 1994

  74. [74]

    Comparison of Proportional Hazards Models and Neural Networks for Reliability Estimation

    James T. Luxhoj and Huan-Jyh Shyur. “Comparison of Proportional Hazards Models and Neural Networks for Reliability Estimation”. In: Journal of Intelligent Manufacturing 8.3 (May 1, 1997), pp. 227–234. ISSN : 1572-8145. DOI: 10.1023/A:1018525308809. 13

    work page doi:10.1023/a:1018525308809 1997

  75. [75]

    Relative survival analysis in R

    M. Pohar and J. Stare. “Relative survival analysis in R”. In:Computer methods and programs in biomedicine 81 (3 2006), pp. 272–278. DOI: 10.1016/j.cmpb.2006.01.004

    work page doi:10.1016/j.cmpb.2006.01.004 2006

  76. [76]

    Evaluating Random Forests for Survival Analysis using Prediction Error Curves

    Ulla B Mogensen, Hemant Ishwaran, and Thomas A Gerds. Evaluating Random Forests for Survival Analysis using Prediction Error Curves. 2014

    work page 2014

  77. [77]

    Statistical Comparison of Survival Models for Analysis of Cancer Data

    Bijan Moghimi-Dehkordi et al. “Statistical Comparison of Survival Models for Analysis of Cancer Data”. In:Asian Pacific journal of cancer prevention: APJCP9.3 (2008), pp. 417–420. ISSN : 2476-762X. pmid: 18990013

    work page 2008

  78. [78]

    General Semiparametric Shared Frailty Model: Estimation and Simulation with frailtySurv

    John V . Monaco, Malka Gorfine, and Li Hsu. “General Semiparametric Shared Frailty Model: Estimation and Simulation with frailtySurv”. In: Journal of Statistical Software 86.4 (2018), pp. 1–42. DOI: 10.18637/jss.v086.i04

    work page doi:10.18637/jss.v086.i04 2018

  79. [79]

    A Comparison of Cox Proportional Hazards and Artificial Neural Network Models for Medical Prognosis

    Lucila Ohno-Machado. “A Comparison of Cox Proportional Hazards and Artificial Neural Network Models for Medical Prognosis”. In:Computers in Biology and Medicine27.1 (Jan. 1, 1997), pp. 55–65. ISSN : 0010-4825. DOI: 10.1016/S0010-4825(96)00036-4

    work page doi:10.1016/s0010-4825(96)00036-4 1997

  80. [80]

    Modeling medical prognosis: survival analysis techniques

    Lucila Ohno-Machado. “Modeling medical prognosis: survival analysis techniques.” eng. In: Journal of biomedical informatics 34.6 (Dec. 2001), pp. 428–439. ISSN : 1532-0464 (Print). DOI: 10.1006/jbin.2002.1038

    work page doi:10.1006/jbin.2002.1038 2001

Showing first 80 references.