Pith. sign in

REVIEW 4 major objections 6 minor 64 references

Survival Analysis as Imprecise Classification with Trainable Kernels

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that survival analysis can be reframed as an imprecise classification problem over time intervals, encoding censored events as interval-valued probabilities and using trainable attention kernels, and that the resulting…

desk verdict A plausible new estimator family for nonparametric survival analysis, but the headline empirical claim is contradicted by the paper's own tables and needs proper benchmarking before being taken seriously. read the letter →

arxiv 2506.10140 v1 pith:4FBLQNNX submitted 2025-06-11 cs.LG stat.ML

Andrei V. Konstantinov , Vlada A. Efremenko , Lev V. Utkin This is my paper · ORCID
classification cs.LGstat.ML
keywords survivalanalysisimpreciseprobabilitiescensoredobservationsattentionmechanismNadaraya-Watsonregressiondiscrete-timenonparametricestimationtrainablekernels
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to establish that time-to-event prediction can be treated as a multiclass classification problem over discrete time intervals, with censored observations represented not by a single distribution but by an interval-valued set of possible distributions. The authors propose three models — iSurvM, iSurvQ, and iSurvJ — that use Nadaraya-Watson regression with trainable attention weights as kernels, so no parametric survival model is assumed. They claim that iSurvJ and its Gaussian-kernel variant iSurvJ(G) consistently outperform the Beran estimator, a classical kernel-based nonparametric estimator of conditional survival, on both accuracy metrics and computational cost, with the advantage growing as feature dimension and censoring fraction increase. If correct, this gives a nonparametric survival method whose smoothing is learned from data rather than fixed, and whose predictions carry interval-valued uncertainty.

What carries the argument

The central object is the interval-valued probability vector $\pi^{(i)}$ over $T$ time intervals: a one-hot vector for uncensored examples and a vector with entries free in $[0,1]$ after the censoring interval for censored examples. The mechanism that carries the argument is a trainable Nadaraya-Watson weighting $p_k(x_0)=\sum_i a_{0,i}(w)\pi^{(i)}_k$, with attention masks preventing self-attention and dot-product or Gaussian kernels providing the weights. Training is driven by likelihood-style losses, and for iSurvJ the loss combines negative log-likelihood with entropy regularization on the learned interval probabilities. To avoid zero gradients from the self-attention mask when all uncensored event times are unique, the paper replaces the exact interval probability with a sum over $2k+1$ neighboring intervals, treating $k$ as a hyperparameter.

What would settle it

Train iSurvJ on a real dataset with tied uncensored event times, such as the BLCD or Rossi data, using the exact-interval loss (Eq. (28), $k=0$) instead of the windowed surrogate (Eq. (29)); if the C-index and integrated Brier score stay competitive with $k>0$, the paper's rationale for the surrogate collapses and the reported comparisons are partly driven by the modified target rather than by the imprecise representation. A more direct check: on a dataset with few ties, compare models trained with Eq. (28) and Eq. (29) and see whether the accuracy gap the paper attributes to the method disappears.

Watch

Extended reading notes

Core claim

The paper's central claim is that right-censored survival data can be encoded as imprecise multi-label classification: an uncensored event contributes a one-hot probability vector over time intervals, while a censored observation contributes the set of all probability vectors whose mass on intervals after the censoring time is free in $[0,1]$. Predictions are formed by the attention-weighted sum $p_k(x_0)=\sum_{i=1}^N a_{0,i}(w)\pi^{(i)}_k$, where the weights come from dot-product attention or a trainable Gaussian kernel. Three training strategies are proposed: averaging likelihood losses over random draws from the censored distributions (iSurvM), averaging only the worst quantile of those draws (iSurvQ), and jointly optimizing the interval probabilities and attention weights with an entropy regularizer (iSurvJ). The paper reports that iSurvJ and iSurvJ(G) beat the Beran estimator on most real and synthetic benchmarks, that the Beran survival curve lies inside the interval-valued survival bounds the model produces, and that the model's unconditional survival function tracks the classical nonparametric estimator.

Load-bearing premise

The load-bearing premise is that training uncensored examples on the sum of probabilities over a window of $2k+1$ neighboring intervals, instead of on the exact probability of the true event interval, is a faithful learning target; the paper justifies this replacement only for datasets with unique uncensored event times and applies it to all datasets without analyzing the bias it may introduce for tied event times or censored data.

Editorial extensions

If this is right

  • The framework removes the need for parametric assumptions: the same training procedure handles linear, nonlinear, and interacting feature structures by learning attention weights from data.
  • Heavy censoring no longer degrades predictions as sharply as it does for the Beran estimator, so the model is a candidate for datasets where most events are unobserved.
  • The interval-valued survival function produced for each instance gives lower and upper bounds that can enclose the Beran estimate, offering uncertainty information alongside point predictions.
  • Because the number of trainable attention parameters can be chosen freely, users can trade model capacity against computational cost without changing the formulation.
  • Among the model's hyperparameters, the window width $k$ for uncensored intervals shows a plateau in accuracy beyond a threshold, so a moderate $k$ avoids unnecessary computation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension the authors leave implicit is to use the interval-valued survival bounds as a formal uncertainty region for decision rules, such as delaying an intervention while the upper survival bound stays high.
  • The same imprecise-classification encoding could be reused for interval-censored and left-censored events by changing which entries of $\pi^{(i)}$ are free, which the paper mentions only briefly as a motivation for imprecision.
  • The $2k+1$ window surrogate can be viewed as label smoothing over neighboring time intervals; if that is the active ingredient, the method's gains might partly come from regularization rather than from the imprecise representation, a hypothesis the paper does not test.
  • Extending the models to competing risks or time-varying covariates, which the authors list as future work, would require replacing the scalar interval target with a vector-valued one but would keep the attention machinery unchanged.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper introduces three survival models, iSurvM, iSurvQ, and iSurvJ (plus a Gaussian-kernel variant iSurvJ(G)), built on a discrete-time, imprecise-probability view of censored survival data. Censored observations are represented as interval-valued probability distributions over time intervals, and predictions are obtained through Nadaraya-Watson-style attention weights trained by one of three loss strategies: mean over Monte Carlo-generated distributions, quantile selection of worst-case losses, or joint learning of probabilities and attention. The central claim is that these models, especially iSurvJ and iSurvJ(G), 'consistently outperform' the Beran estimator in both accuracy and computational complexity. The paper reports C-index and Brier-score comparisons on 11 real datasets and several synthetic benchmarks, and includes public code.

Significance. If the claims were established, the paper would contribute a novel synthesis of imprecise probability and attention-based kernel smoothing for survival analysis, with no parametric assumptions and with an explicit interval-valued representation of censoring uncertainty. The availability of public code and the breadth of datasets are positive features. However, the headline empirical claim is not supported by the paper's own tables: iSurvJ loses to Beran on several datasets, iSurvJ(G) loses on a substantial fraction of comparisons, and no uncertainty estimates or significance tests are provided. The computational-complexity claim is asserted rather than measured. Because the central contribution is framed as consistent superiority over Beran, the current evidence is insufficient for the stated conclusions.

major comments (4)
  1. [Section 6, Tables 1-2] The abstract and Section 6 claim that the proposed models 'consistently outperform' the Beran estimator, but Tables 1 and 2 show multiple losses. For example, in Table 1 iSurvJ has a lower C-index than Beran on AIDS (0.7139 vs. 0.7529) and on Breast Cancer (0.6487 vs. 0.6519), while iSurvJ(G) is worse than Beran on Veterans, AIDS, GBSG2, LND, and GCD. In Table 2, iSurvJ(G) has a worse Brier score than Beran on 6 of 11 datasets, and iSurvJ is worse on AIDS, Breast Cancer, and LND. No standard deviations, confidence intervals, or significance tests are reported, so even the apparent wins cannot be distinguished from noise, particularly on small datasets such as LND (n=87) and GCD (n=90). The accuracy half of the headline claim is therefore not supported by the presented evidence.
  2. [Section 5.5, Eq. (29)] The replacement of the exact event-interval loss by a sum over 2k+1 neighboring intervals is justified by the argument that, when all event times are unique and uncensored, the self-attention mask removes the only training example that can contribute to p_c(i), making the exact loss zero. That argument does not apply to datasets containing censored observations or tied event intervals, where other training examples contribute mass to the correct-interval probability. The modified loss is nevertheless applied to all models and all datasets, and no analysis of the resulting bias is provided. Since this surrogate loss is used in every accuracy comparison, the experimental results rest on a learning signal whose validity is not established for the data actually used.
  3. [Section 6, Paragraph on hyperparameter tuning] The comparison between the proposed models and the Beran estimator is asymmetric. The proposed models receive Optuna-based Bayesian hyperparameter optimization over embedding dimensions, learning rates, regularization coefficients, dropout, mask rates, batch sizes, and epoch counts, whereas no comparable tuning procedure for the Beran estimator is described; the only Beran parameter mentioned is a Gaussian kernel temperature in Appendix A.3.1. Observed differences could therefore reflect differences in tuning effort rather than a substantive advantage of the proposed models. The authors should either tune the Beran estimator with the same budget or justify why the chosen baseline configuration is the appropriate reference.
  4. [Section 6 and Appendix A.3.4] The claim of superiority 'from the computational complexity point of view' is asserted without any runtime measurements or asymptotic analysis; no wall-clock times, flop counts, or scaling experiments are reported. In addition, the interval bounds in Appendix A.3.4 are not strong evidence for the model: because the censored interval probabilities are allowed to range over [0,1] by construction, the resulting survival-function bounds are very wide and any reasonable survival curve will lie inside them. The observation that Beran's curve falls inside the bounds should be presented as an illustration of the imprecise representation, not as an accuracy or validity finding.
minor comments (6)
  1. [Section 1, Contributions list] In the contributions list, the second model is written as 'iSurvM' twice: the text says 'survival models iSurvM, iSurvM, iSurvM' where the second and third should likely be iSurvQ and iSurvJ.
  2. [Algorithm 2 caption] The caption of Algorithm 2 says 'An implementation of training the model iSurvM' but the algorithm is for iSurvJ; this should be corrected.
  3. [Equation (2)] The C-index formula uses predicted expected event times \widehat{T}_i and \widehat{T}_j, but how these expected times are computed from the discrete predicted distribution is not defined anywhere in the paper.
  4. [Equation (7)] The Gaussian attention weight formula has a notational issue: the parameter vector is denoted w = (w_1, ..., w_n), but the denominator uses the same w_i for every term, which cannot be correct unless all weights are equal; the dependence on the target feature index should be made explicit.
  5. [Appendix A.3.3] The text says the experiments 'primarily illustrate how the accuracy of the proposed model decreases as the proportion of censored data increases,' yet the abstract and Section 6 claim robustness to heavy censoring; the two statements should be reconciled.
  6. [Figures 8-9 captions] The captions contain the typo 'censiring rate' instead of 'censoring rate'.

Circularity Check

1 steps flagged · score 2.0 of 10

Central empirical comparison is independent of Beran, but the claim that Beran's SF lies inside the model's interval bounds is true largely by construction.

  1. self definitional [Appendix A.3.4, 'Intervals for Survival Functions'; also referenced in Section 6]
    "The bounds for the SF are obtained using (12), where probabilities π(i)_k, k = 1, ..., T, are interval-valued for censored observations. ... It is important to note that the Beran estimator predicts the SF which is totally inside the lower and upper SF bounds. This is a very interesting observation."

    By Eq. (11), every censored observation assigns probability [0,1] to each interval after its censoring time. Eq. (12) then forms the predicted interval probabilities as convex combinations of these interval values, so the resulting SF bounds are extremely wide envelopes rather than informative model-specific limits. Any survival curve whose probabilities over the partition are consistent with the censoring pattern — including the Beran estimate — will fall inside such bounds almost automatically. The containment is thus a consequence of the interval construction, not an independent validation that the imprecise model 'encapsulates' the Beran estimator.

full rationale

The paper's main claim is an empirical comparison: iSurvJ and iSurvJ(G) are trained on survival data and evaluated with the C-index and Brier score against the Beran estimator on held-out folds. That comparison is not circular: the model outputs are not constructed from Beran's predictions, and the Beran baseline is an external estimator. No load-bearing self-citation or imported uniqueness theorem appears; the cited kernel/attention results are standard external tools. The only near-circular element is the interval-bounds observation in Appendix A.3.4, where the claim that the Beran SF lies inside the model's lower and upper SF bounds is true largely by construction because each censored interval contributes [0,1] to the interval-valued probabilities, making the envelope very wide. That observation is presented as supporting evidence for the model's uncertainty representation, but it is weak evidence and should not be read as an independent confirmation. Since the central accuracy claim stands or falls on the external experiments rather than on this containment argument, the overall circularity is minor: one structurally trivial containment step, not a derivation that reduces to its own inputs.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

All three models rely on assumptions introduced without proof: the time discretization, the interval-valued representation of censoring, Dirichlet sampling as a proxy for interval ignorance, the attention-weighted average as a valid estimator, and a widened target for uncensored events. The trainable parameters of the attention network are fitted on data, but many structural hyperparameters (k, M, r, gamma, pmask, T) are chosen by hand or by Optuna and are not derived.

free parameters (7)
  • k, neighborhood window in Eq. 29 = tuned between 3 and 10
    Replaces the exact interval loss for uncensored instances; introduced to avoid zero gradients from self-masking. The plateau behavior in Fig. 3 shows it is a trade-off hyperparameter, not a derived quantity.
  • M, number of Monte Carlo generations = not stated in the paper
    iSurvM and iSurvQ average the loss over M Dirichlet samples; the choice of M is a hyperparameter and is not specified.
  • r, quantile fraction in iSurvQ = not stated in the paper
    Selects the r*M largest losses in Eq. 26; the paper gives no guidance or sensitivity analysis.
  • gamma, entropy regularization weight = 0.1 in appendix experiments
    Appears in Eq. 27 and in fine-tuning; tuned via Optuna, no derivation.
  • pmask, attention mask threshold = 0.1 to 0.5 in hyperparameter search
    The random mask in Eq. 16 is binarized with this threshold; an arbitrary architectural choice.
  • T and interval boundaries = not specified
    The partition in Eq. 8 determines all probability vectors; the paper does not state how T or the boundaries are chosen for each dataset.
  • Dirichlet concentration parameters = not specified
    Section 5.1 says distributions are generated via the Dirichlet distribution, but no concentration parameters are given; this determines the sampled distributions for censored observations.
assumptions (6)
  • domain assumption The time axis partition into T intervals (Eq. 8) is fixed and each event time is represented by exactly one interval.
    All probability and loss definitions in Section 4 depend on this discretization; the paper does not specify how T or the boundaries are chosen for each dataset.
  • domain assumption For censored observations, the probability of each future interval is in [0,1] (Eq. 11), expressing prior ignorance.
    This interval-valued representation is the entire basis for the imprecise classification framework; no justification beyond 'a form of prior ignorance' is given.
  • ad hoc to paper Sampling censored probability distributions via the Dirichlet distribution correctly represents the set R(i).
    Section 5.1 introduces Dirichlet generation with no concentration parameters and no theoretical link between the [0,1] intervals and the Dirichlet draws.
  • domain assumption The attention-weighted average p_k(x0) = sum_i a_{0,i}(w) pi_i(k) (Eq. 12) is a valid estimator of class probabilities for a new instance.
    Borrowed from Nadaraya-Watson regression; valid under standard smoothing assumptions, but no consistency result is stated for interval-valued labels.
  • ad hoc to paper The extended-interval loss (Eq. 29) is a valid surrogate for exact interval prediction for uncensored data.
    Introduced to avoid zero gradients from self-masking; it widens the target to 2k+1 intervals and is applied to all models. Its bias is not analyzed.
  • ad hoc to paper Attention masking with a random threshold pmask (Eq. 16) does not distort learning.
    The mask is random, binarized, and diagonal-excluded; no justification or ablation is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Survival Analysis as Imprecise Classification with Trainable Kernels." pith.science (2026). https://pith.science/paper/4FBLQNNX

@misc{pith2026250610140,
  author       = {Pith},
  title        = {Pith review of: Survival Analysis as Imprecise Classification with Trainable Kernels},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4FBLQNNX}},
  note         = {Machine review of arXiv:2506.10140}
}
read the original abstract

Survival analysis is a fundamental tool for modeling time-to-event data in healthcare, engineering, and finance, where censored observations pose significant challenges. While traditional methods like the Beran estimator offer nonparametric solutions, they often struggle with the complex data structures and heavy censoring. This paper introduces three novel survival models, iSurvM (the imprecise Survival model based on Mean likelihood functions), iSurvQ (the imprecise Survival model based on the Quantiles of likelihood functions), and iSurvJ (the imprecise Survival model based on the Joint learning), that combine imprecise probability theory with attention mechanisms to handle censored data without parametric assumptions. The first idea behind the models is to represent censored observations by interval-valued probability distributions for each instance over time intervals between events moments. The second idea is to employ the kernel-based Nadaraya-Watson regression with trainable attention weights for computing the imprecise probability distribution over time intervals for the entire dataset. The third idea is to consider three decision strategies for training, which correspond to the proposed three models. Experiments on synthetic and real datasets demonstrate that the proposed models, especially iSurvJ, consistently outperform the Beran estimator from the accuracy and computational complexity points of view. Codes implementing the proposed models are publicly available.

Figures

Figures reproduced from arXiv: 2506.10140 by the authors.

Figure 1
Figure 1. Dependence of the C-index and IBS on the number of features for the Linear, Quadratic, [PITH_FULL_IMAGE:figures/full_fig_p023_1.png] view at source ↗
Figure 2
Figure 2. Dependence of the C-index and IBS on the number of features for the Sparse Features, [PITH_FULL_IMAGE:figures/full_fig_p024_2.png] view at source ↗
Figure 3
Figure 3. Dependence of the C-index and IBS on the parameter [PITH_FULL_IMAGE:figures/full_fig_p026_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Dependence of the C-index and IBS on the proportion of censored data for the Strong Feature [PITH_FULL_IMAGE:figures/full_fig_p027_4.png]
Figure 5
Figure 5. Figure 5: Comparison of iSurvJ(G) and the Beran estimator by different values of the censored obser [PITH_FULL_IMAGE:figures/full_fig_p028_5.png]
Figure 6
Figure 6. Figure 6: Comparison of iSurvJ(G) and the Beran estimator by different values of the censored obser [PITH_FULL_IMAGE:figures/full_fig_p029_6.png]
Figure 7
Figure 7. Figure 7: Comparison of iSurvJ(G) and the Beran estimator by different values of the censored obser [PITH_FULL_IMAGE:figures/full_fig_p030_7.png]
Figure 8
Figure 8. Figure 8: Interval-valued SFs for the Friedman1 dataset by different values of the censiring rate [PITH_FULL_IMAGE:figures/full_fig_p031_8.png]
Figure 9
Figure 9. Figure 9: Interval-valued SFs for the Nonlinear dataset by different values of the censiring rate [PITH_FULL_IMAGE:figures/full_fig_p032_9.png]
Figure 10
Figure 10. Figure 10: Unconditional SFs for the Veterans and GBSG2 datasets [PITH_FULL_IMAGE:figures/full_fig_p033_10.png]
Figure 11
Figure 11. Figure 11: Unconditional SFs for the WHAS500 and Breast Cancer datasets [PITH_FULL_IMAGE:figures/full_fig_p034_11.png]
Figure 12
Figure 12. Figure 12: Comparison of predicted expected times obtained by the Beran estimator and iSurvJ(G) [PITH_FULL_IMAGE:figures/full_fig_p035_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 63 canonical work pages

  1. [1]

    Hosmer, S

    D. Hosmer, S. Lemeshow, and S. May. Applied Survival Analysis: Regression Modeling of Time to Event Data . John Wiley & Sons, New Jersey, 2008

    work page 2008

  2. [2]

    A deep survival analysis method based on ranking

    Bingzhong Jing, Tao Zhang, Zixian Wang, Ying Jin, Kuiyuan Liu, Wenze Qiu, Liangru Ke, Ying Sun, Caisheng He, Dan Hou, et al. A deep survival analysis method based on ranking. Artificial intelligence in medicine , 98:1–9, 2019

    work page 2019

  3. [3]

    Deep learning-based survival analysis for high-dimensional survival data

    Lin Hao, Juncheol Kim, Sookhee Kwon, and Il Do Ha. Deep learning-based survival analysis for high-dimensional survival data. Mathematics, 9(11):1244, 2021

    work page 2021

  4. [4]

    Lee and J.W

    E.T. Lee and J.W. Wang. Statistical Methods for Survival Data Analysis . John Wiley & Sons, New Jersey, 2003

    work page 2003

  5. [5]

    Hothorn, P

    T. Hothorn, P. B¨ uhlmann, S. Dudoit, A. Molinaro, and M.J. van der Laan. Survival ensembles. Biostatistics, 7(3):355–373, 2006

    work page 2006

  6. [6]

    Wrobel, A

    L. Wrobel, A. Gudys, and M. Sikora. Learning rule sets from survival data. BMC Bioinformatics, 18(1):285–297, 2017

    work page 2017

  7. [7]

    DNNSurv: Deep Neural Networks for Survival Analysis Using Pseudo Values

    L. Zhao and D. Feng. Dnnsurv: Deep neural networks for survival analysis using pseudo values. arXiv:1908.02337v2, March 2020. 14

    work page Pith review arXiv 1908

  8. [8]

    Tutorial on survival modeling with applications to omics data

    Zhi Zhao, John Zobolas, Manuela Zucknick, and Tero Aittokallio. Tutorial on survival modeling with applications to omics data. Bioinformatics, 40(3):btae132, 2024

    work page 2024

Show all 64 references
  1. [9]

    Marinos and D

    G. Marinos and D. Kyriazis. A survey of survival analysis techniques. In HEALTHINF, pages 716–723, 2021

    2021

  2. [10]

    P. Wang, Y. Li, and C.K. Reddy. Machine learning for survival analysis: A survey.ACM Computing Surveys (CSUR) , 51(6):1–36, 2019

    2019

  3. [11]

    Deep learning for survival analysis: a review

    Simon Wiegrebe, Philipp Kopper, Raphael Sonabend, Bernd Bischl, and Andreas Bender. Deep learning for survival analysis: a review. Artificial Intelligence Review , 57(65):1–34, 2024

    2024

  4. [12]

    Ishwaran and U.B

    H. Ishwaran and U.B. Kogalur. Random survival forests for r. R News , 7(2):25–31, 2007

    2007

  5. [13]

    Van Belle, K

    V. Van Belle, K. Pelckmans, J.A.K. Suykens, and S. Van Huffel. Survival svm: a practical scalable algorithm. In ESANN, pages 89–94, 2008

    2008

  6. [14]

    An introduction to deep survival analysis models for predicting time-to-event outcomes

    George H Chen et al. An introduction to deep survival analysis models for predicting time-to-event outcomes. Foundations and Trends® in Machine Learning , 17(6):921–1100, 2024

    2024

  7. [15]

    Arroyo, A

    A. Arroyo, A. Cartea, F. Moreno-Pino, and S. Zohren. Deep attentive survival analysis in limit order books: Estimating fill probabilities with convolutional-transformers. Quantitative Finance, 24(1):35–57, 2024

    2024

  8. [16]

    Transformer-based deep survival analysis

    Shi Hu, Egill Fridgeirsson, Guido van Wingen, and Max Welling. Transformer-based deep survival analysis. In Survival Prediction-Algorithms, Challenges and Applications , pages 132–148. PMLR, 2021

    2021

  9. [17]

    Hierarchical transformer for sur- vival prediction using multimodality whole slide images and genomics

    Chunyuan Li, Xinliang Zhu, Jiawen Yao, and Junzhou Huang. Hierarchical transformer for sur- vival prediction using multimodality whole slide images and genomics. In The 26th International Conference on Pattern Recognition (ICPR) , pages 4256–4262. IEEE Computer Society, 2022

    2022

  10. [18]

    Adaptive transformer modelling of density function for nonparametric survival analysis

    Xin Zhang, Deval Mehta, Yanan Hu, Chao Zhu, David Darby, Zhen Yu, Daniel Merlo, Melissa Gresle, Anneke Van Der Walt, Helmut Butzkueven, et al. Adaptive transformer modelling of density function for nonparametric survival analysis. Machine Learning, 114(2):31, 2025

    2025

  11. [19]

    Kaplan and P

    E.L. Kaplan and P. Meier. Nonparametric estimation from incomplete observations. Journal of the American Statistical Association , 53(282):457–481, 1958

    1958

  12. [20]

    R. Beran. Nonparametric regression with randomly censored survival data. Technical report, University of California, Berkeley, 1981

    1981

  13. [21]

    Survival kernets: Scalable and interpretable deep kernel survival analysis with an accuracy guarantee

    George H Chen. Survival kernets: Scalable and interpretable deep kernel survival analysis with an accuracy guarantee. Journal of Machine Learning Research , 25(40):1–78, 2024

    2024

  14. [22]

    Sparse kernel methods for high-dimensional survival data

    Ludger Evers and Claudia-Martina Messow. Sparse kernel methods for high-dimensional survival data. Bioinformatics, 24(14):1632–1638, 2008

    2008

  15. [23]

    A review on smoothing methods for the estimation of the hazard rate based on kernel functions

    O Gefeller and P Michels. A review on smoothing methods for the estimation of the hazard rate based on kernel functions. In Computational Statistics: Volume 1: Proceedings of the 10th Symposium on Computational Statistics , pages 459–464. Springer, 1992. 15

    1992

  16. [24]

    Cawley, N.L.C

    G.C. Cawley, N.L.C. Talbot, G.J. Janacek, and M.W. Peck. Bayesian kernel learning methods for parametric accelerated life survival analysis. In Proceedings of the First international conference on Deterministic and Statistical Methods in Machine Learning , pages 37–55, 2004

    2004

  17. [25]

    Kernel cox regression models for linking gene expression profiles to censored survival data

    Hongzhe Li and Yihui Luan. Kernel cox regression models for linking gene expression profiles to censored survival data. In Biocomputing 2003, pages 65–76. World Scientific, 2002

    2003

  18. [26]

    Kernel cox partially linear regression: Building predictive models for cancer patients’ survival

    Yaohua Rong, Sihai Dave Zhao, Xia Zheng, and Yi Li. Kernel cox partially linear regression: Building predictive models for cancer patients’ survival. Statistics in Medicine , 43(1):1–15, 2024

    2024

  19. [27]

    Weighted functional linear cox regression model

    Hojin Yang, Hongtu Zhu, Mihye Ahn, and Joseph G Ibrahim. Weighted functional linear cox regression model. Statistical Methods in Medical Research, 30(8):1917–1931, 2021

    1917

  20. [28]

    D.R. Cox. Regression models and life-tables. Journal of the Royal Statistical Society, Series B (Methodological), 34(2):187–220, 1972

    1972

  21. [29]

    Tutz and M

    G. Tutz and M. Schmid. Modeling discrete time-to-event data . Springer, 2016

    2016

  22. [30]

    Survival prediction models: an introduc- tion to discrete-time modeling

    Krithika Suresh, Cameron Severn, and Debashis Ghosh. Survival prediction models: an introduc- tion to discrete-time modeling. BMC medical research methodology, 22(1):207, 2022

    2022

  23. [31]

    Continuous and discrete-time survival prediction with neural networks

    H ˚ avard Kvamme and Ørnulf Borgan. Continuous and discrete-time survival prediction with neural networks. Lifetime data analysis , 27(4):710–736, 2021

    2021

  24. [32]

    Survival analysis as a classification problem

    C Zhong and R Tibshirani. Survival analysis as a classification problem. arXiv:1909.11171v2, Sep 2019

    1909 arXiv

  25. [33]

    Nadaraya

    E.A. Nadaraya. On estimating regression. Theory of Probability & Its Applications , 9(1):141–142, 1964

    1964

  26. [34]

    G.S. Watson. Smooth regression analysis. Sankhya: The Indian Journal of Statistics, Series A , pages 359–372, 1964

    1964

  27. [35]

    Luong, H

    T. Luong, H. Pham, and C.D. Manning. Effective approaches to attention-based neural machine translation. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 1412–1421. The Association for Computational Linguistics, 2015

    2015

  28. [36]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A.N. Gomez, L. Kaiser, and I. Polo- sukhin. Attention is all you need. In Advances in Neural Information Processing Systems , pages 5998–6008, 2017

    2017

  29. [37]

    Kvamme, O

    H. Kvamme, O. Borgan, and I. Scheel. Time-to-event prediction with neural networks and cox regression. Journal of Machine Learning Research , 20:1–30, 2019

    2019

  30. [38]

    F.P.A. Coolen. An imprecise Dirichlet model for Bayesian analysis of failure data including right- censored observations. Reliability Engineering and System Safety , 56:61–68, 1997

    1997

  31. [39]

    Coolen and K.J

    F.P.A. Coolen and K.J. Yan. Nonparametric predictive inference withright-censored data. Journal of Statistical Planning andInference , 126:25–54, 2004

    2004

  32. [40]

    Mangili, A

    F. Mangili, A. Benavoli, C.P. de Campos, and M. Zaffalon. Reliable survival analysis based on the Dirichlet process. Biometrical Journal, 57(6):1002–1019, 2015. 16

    2015

  33. [41]

    Explainable survival analysis with uncertainty using convolution- involved vision transformer

    Zhihao Tang, Li Liu, Zongyi Chen, Guixiang Ma, Jiyan Dong, Xujie Sun, Xi Zhang, Chaozhuo Li, Qingfeng Zheng, Lin Yang, et al. Explainable survival analysis with uncertainty using convolution- involved vision transformer. Computerized Medical Imaging and Graphics , 110:102302, 2023

    2023

  34. [42]

    Resdeepsurv: A survival model for deep neural networks based on residual blocks and self-attention mechanism

    Yuchen Wang, Xianchun Kong, Xiao Bi, Lizhen Cui, Hong Yu, and Hao Wu. Resdeepsurv: A survival model for deep neural networks based on residual blocks and self-attention mechanism. Interdisciplinary Sciences: Computational Life Sciences , 16(2):405–417, 2024

    2024

  35. [43]

    High-dimensional survival analysis: Methods and applications

    Stephen Salerno and Yi Li. High-dimensional survival analysis: Methods and applications. Annual review of statistics and its application , 10:25–49, 2023

    2023

  36. [44]

    A general machine learning framework for survival analysis

    Andreas Bender, David R¨ ugamer, Fabian Scheipl, and Bernd Bischl. A general machine learning framework for survival analysis. In Joint European conference on machine learning and knowledge discovery in databases , pages 158–173. Springer, 2020

    2020

  37. [45]

    Emmert-Streib and M

    F. Emmert-Streib and M. Dehmer. Introduction to survival analysis in practice. Machine Learning & Knowledge Extraction , 1:1013–1038, 2019

    2019

  38. [46]

    Deep kernel survival analysis and subject-specific survival time prediction intervals

    George H Chen. Deep kernel survival analysis and subject-specific survival time prediction intervals. In Machine learning for healthcare conference, pages 537–565. PMLR, 2020

    2020

  39. [47]

    Deep gated neural network with self-attention mechanism for survival analysis

    Xulin Yang and Hang Qiu. Deep gated neural network with self-attention mechanism for survival analysis. IEEE Journal of Biomedical and Health Informatics , 29(4):2945–2956, 2024

    2024

  40. [48]

    Survtrace: Transformers for survival analysis with competing events

    Zifeng Wang and Jimeng Sun. Survtrace: Transformers for survival analysis with competing events. In Proceedings of the 13th ACM International Conference on Bioinformatics, Computational Biol- ogy and Health Informatics , pages 1–9, 2022

    2022

  41. [49]

    Mhattnsurv: Multi-head attention for survival prediction using whole-slide pathology images

    Shuai Jiang, Arief A Suriawinata, and Saeed Hassanpour. Mhattnsurv: Multi-head attention for survival prediction using whole-slide pathology images. Computers in biology and medicine , 158:106883, 2023

    2023

  42. [50]

    A semi-supervised transformer survival prediction model for lung cancer

    Jing Teng, Lan Yang, Shan Wang, and Jing Yu. A semi-supervised transformer survival prediction model for lung cancer. Advanced Functional Materials, page 2419005, 2025

    2025

  43. [51]

    Surformer: An interpretable pattern-perceptive survival transformer for cancer survival prediction from histopathology whole slide images

    Zhikang Wang, Qian Gao, Xiaoping Yi, Xinyu Zhang, Yiwen Zhang, Daokun Zhang, Pietro Li` o, Chris Bain, Richard Bassed, Shanshan Li, et al. Surformer: An interpretable pattern-perceptive survival transformer for cancer survival prediction from histopathology whole slide images....

    2023

  44. [52]

    A multi-head attention trans- former framework for oesophageal cancer survival prediction

    Zhongyu Yao, Tianhang Chen, Lingkuan Meng, and Ka-Chun Wong. A multi-head attention trans- former framework for oesophageal cancer survival prediction. In 2024 4th International Conference on Artificial Intelligence, Robotics, and Communication (ICAIRC) , pages 309–313. IEEE, 2024

    2024

  45. [53]

    De finetti’s theorem, induction, and a (n) or bayesian nonparametric predictive inference (with discussion)

    Bruce M Hill. De finetti’s theorem, induction, and a (n) or bayesian nonparametric predictive inference (with discussion). Bayesian statistics , 3:211–241, 1988

    1988

  46. [54]

    P. Walley. Inferences from multinomial data: Learning about a bag of marbles. Journal of the Royal Statistical Society, Series B , 58:3–57, 1996. with discussion

    1996

  47. [55]

    Harrell, R

    F. Harrell, R. Califf, D. Pryor, K. Lee, and R. Rosati. Evaluating the yield of medical tests. Journal of the American Medical Association , 247:2543–2546, 1982. 17

    1982

  48. [56]

    M. May, P. Royston, M. Egger, A.C. Justice, and J.A.C. Sterne. Development and validation of a prognostic model for survival time data: application to prognosis of HIV positive patients treated with antiretroviral therapy. Statistics in Medicine , 23:2375–2398, 2004

    2004

  49. [57]

    Uno, Tianxi Cai, M.J

    H. Uno, Tianxi Cai, M.J. Pencina, R.B. D’Agostino, and Lee-Jen Wei. On the c-statistics for evaluating overall adequacy of risk prediction procedures with censored survival data. Statistics in medicine, 30(10):1105–1117, 2011

    2011

  50. [58]

    G.W. Brier. Verification of forecasts expressed in terms of probability. Monthly Weather Review , 78(1):1–3, 1950

    1950

  51. [59]

    E. Graf, C. Schmoor, W. Sauerbrei, and M. Schumacher. Assessment and comparison of prognostic classification schemes for survival data. Statistics in Medicine , 18(17-18):2529–2545, 1999

    1999

  52. [60]

    Bahdanau, K

    D. Bahdanau, K. Cho, and Y. Bengio. Neural machine translation by jointly learning to align and translate. arXiv:1409.0473, Sep 2014

    2014 arXiv

  53. [61]

    Rubinstein and D.P

    R.Y. Rubinstein and D.P. Kroese. Simulation and the Monte Carlo method, 2nd Edition . Wiley, New Jersey, 2008

    2008

  54. [62]

    Smith and R.W

    N.A. Smith and R.W. Tromble. Sampling uniformly from the unit simplex. Technical Report 29, Johns Hopkins University, 2004

    2004

  55. [63]

    Akiba, S

    T. Akiba, S. Sano, T. Yanase, T. Ohta, and M. Koyama. Optuna: A next-generation hyperparam- eter optimization framework. In Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining , pages 2623–2631, 2019

    2019

  56. [64]

    survival

    S. Destercke and V. Antoine. Combining imprecise probability masses with maximal coherent subsets: Application to ensemble classification. In Synergies of Soft Computing and Statistics for Intelligent Data Analysis , pages 27–35. Springer, Berlin, Heidelberg, 2013. A Appendix:...

    2013

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.