pith. sign in

arxiv: 2604.26559 · v1 · submitted 2026-04-29 · 📊 stat.ME · math.ST· stat.TH

Principled Estimation and Prediction with Competing Risks: a Bayesian Nonparametric Approach

Claudio Del Sole , Antonio Lijoi , Igor Pr\"unster This is my paper

Pith reviewed 2026-05-07 12:35 UTC · model grok-4.3

classification 📊 stat.ME math.STstat.TH
keywords competing risksBayesian nonparametricprediction curvesurvival analysiscompletely random measurestransition probabilitiesmulti-state models
0
0 comments X

The pith

A Bayesian nonparametric model for competing risks defines a prediction curve that gives the probability a future event is of a specific type as a function of its time.

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

The paper places competing risks inside a multi-state framework and models the transition probabilities with a new class of nonparametric priors built from hierarchical completely random measures. It isolates the conditionally conjugate member of this class, derives the joint distribution of the observed data and a latent random partition, and fully characterizes the posterior. These distributional results directly yield the prediction curve, which reports the conditional probability that an event occurring at a given time belongs to one particular cause. The same posterior also supplies estimates of the overall survival function and the cause-specific incidence and subdistribution functions, together with algorithms that make the computations feasible.

Core claim

The central claim is that priors constructed from hierarchical completely random measures on transition probabilities admit an explicit posterior and a latent partition representation, from which the prediction curve follows as the time-dependent predictive probability that a future event belongs to a given cause.

What carries the argument

The prediction curve, obtained by integrating the posterior distribution of the transition probabilities to give the probability that an event at time t is of a specified type.

If this is right

  • Posterior estimates become available for the survival function, cause-specific incidence functions, and subdistribution functions.
  • Simulation algorithms can be used to draw from the posterior for any observed competing-risks dataset.
  • The prediction curve supplies cause-specific forecasts conditional on the timing of the next event.
  • The same machinery applies directly to clinical data with multiple possible event types.

Where Pith is reading between the lines

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

  • The prediction curve could be computed in actuarial or reliability settings where multiple failure modes compete.
  • Covariates could be incorporated by extending the hierarchical measure construction without changing the overall conjugacy argument.
  • The latent partition representation might simplify model checking by revealing clusters of similar transition behavior across subjects.

Load-bearing premise

The class of priors defined through hierarchical completely random measures is flexible enough to represent transition probabilities in competing risks without forcing restrictive forms.

What would settle it

A held-out dataset in which the model's computed prediction curve deviates systematically from the empirical frequencies of each cause at different times would show that the posterior-derived probabilities are not reliable.

Figures

Figures reproduced from arXiv: 2604.26559 by Antonio Lijoi, Claudio Del Sole, Igor Pr\"unster.

Figure 1
Figure 1. Figure 1: Prediction curves corresponding to melanoma and other causes of death for a view at source ↗
Figure 2
Figure 2. Figure 2: Survival function for a synthetic dataset with view at source ↗
Figure 3
Figure 3. Figure 3: Cause–specific incidence functions (left), subdistributions (middle) and view at source ↗
Figure 4
Figure 4. Figure 4: EBMT dataset: posterior estimates of the survival function (left), the view at source ↗
read the original abstract

Competing risks occur in survival analysis when multiple causes of death are present. They play a prominent role in several domains extending beyond biostatistics to encompass epidemiology, actuarial sciences, and reliability theory. This paper adopts a multi-state modeling framework to competing risks. We introduce a class of flexible nonparametric priors, defined through hierarchical completely random measures, to model the transition probabilities, and identify the specific (conditionally) conjugate member of this general class. Furthermore, we determine the joint marginal distribution of the data and of a latent random partition, and characterize the posterior distribution of the model. Leveraging these distributional results, we evaluate the predictive probability that a future event is of a specific type (e.g. death from a particular cause), as a function of the time at which the event occurs. The resulting function, derived on sound principles, is termed the prediction curve, and represents a major innovation in the literature. In addition, we provide posterior estimates for the survival function, and for the cause-specific incidence and subdistribution functions. Suitable simulation algorithms for posterior inference are also devised. The model's performance, as well as the algorithms' effectiveness, is evaluated through simulation studies. Finally, we illustrate our approach on clinical datasets.

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

2 major / 3 minor

Summary. The manuscript introduces a Bayesian nonparametric model for competing risks in a multi-state framework by employing hierarchical completely random measures as priors for transition probabilities. A conditionally conjugate member of this class is identified, allowing derivation of the joint marginal distribution of the data and latent random partition, as well as characterization of the posterior distribution. Using these results, the authors define the prediction curve, which represents the predictive probability of a future event being of a specific type conditional on the time of occurrence. The paper also derives posterior estimates for the survival function, cause-specific incidence functions, and subdistribution functions, develops simulation algorithms for inference, evaluates the approach through simulation studies, and applies it to clinical datasets.

Significance. Should the central derivations prove correct and the conjugate prior sufficiently flexible, this work offers a principled Bayesian nonparametric framework for estimation and prediction in competing risks settings. The introduction of the prediction curve as a time-dependent predictive tool is a potentially major innovation, providing a new way to assess cause-specific risks over time. The availability of posterior inference for standard survival quantities and the computational algorithms add practical value. Simulation studies and real-data illustration support the method's utility, though their depth determines the strength of evidence.

major comments (2)
  1. [Model specification and prior choice] The identification of the conditionally conjugate member of the hierarchical CRM class (detailed in the model specification) is load-bearing for all subsequent distributional results and the prediction curve. Conjugacy may restrict flexibility in modeling dependencies between cause-specific hazards or tail behaviors of transition probabilities; the manuscript must demonstrate (via theoretical bounds or targeted simulations comparing to the non-conjugate class) that this choice preserves necessary modeling power, as any restriction would propagate directly to the derived predictive probabilities as a function of event time.
  2. [Derivation of the prediction curve] The prediction curve is constructed by leveraging the joint marginal of data and latent partition together with the characterized posterior. The manuscript should provide explicit steps (referencing the relevant equations in the derivation section) showing that this does not reduce by construction to quantities already determined by the fitted model, to substantiate the claim of a major innovation derived on sound principles.
minor comments (3)
  1. [Abstract] The abstract asserts that the prediction curve 'represents a major innovation in the literature' but provides no citations to prior work on time-dependent prediction or cause-specific hazards in competing risks; adding 2-3 targeted references would better position the contribution.
  2. [Simulation studies] In the simulation studies section, include quantitative metrics (e.g., bias, coverage, or integrated error for the prediction curve estimates across scenarios) rather than qualitative statements on performance and algorithm effectiveness.
  3. [Notation and model setup] Clarify the notation for the latent random partition and transition probability matrices early in the model section to ensure consistency when deriving the joint marginal and posterior.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and insightful comments on our manuscript. We address each major comment in detail below, indicating the revisions we will make to improve clarity and strengthen the supporting evidence.

read point-by-point responses

  1. Referee: [Model specification and prior choice] The identification of the conditionally conjugate member of the hierarchical CRM class (detailed in the model specification) is load-bearing for all subsequent distributional results and the prediction curve. Conjugacy may restrict flexibility in modeling dependencies between cause-specific hazards or tail behaviors of transition probabilities; the manuscript must demonstrate (via theoretical bounds or targeted simulations comparing to the non-conjugate class) that this choice preserves necessary modeling power, as any restriction would propagate directly to the derived predictive probabilities as a function of event time.

    Authors: We agree that the conditionally conjugate member is central to obtaining the closed-form distributional results and the prediction curve. While conjugacy necessarily imposes some structure, the specific member we identify within the hierarchical CRM class remains highly flexible: it can accommodate arbitrary dependence structures between cause-specific hazards and a wide range of tail behaviors through the choice of the underlying Lévy measure and hyperparameters. To provide concrete evidence that this choice does not materially restrict modeling power, we will add a new subsection in the simulation studies that compares the conjugate model to a non-conjugate hierarchical CRM (approximated via MCMC) across scenarios with strong cross-cause dependence and heavy-tailed transition probabilities. These comparisons will focus on bias and coverage of the prediction curve and incidence functions. revision: partial

  2. Referee: [Derivation of the prediction curve] The prediction curve is constructed by leveraging the joint marginal of data and latent partition together with the characterized posterior. The manuscript should provide explicit steps (referencing the relevant equations in the derivation section) showing that this does not reduce by construction to quantities already determined by the fitted model, to substantiate the claim of a major innovation derived on sound principles.

    Authors: We thank the referee for this request for greater explicitness. The prediction curve is obtained by first marginalizing the type-specific probability over the posterior distribution of the latent partition and the random transition measures; this integral is not a deterministic function of the fitted posterior means but explicitly accounts for the uncertainty in the partition structure. In the revised manuscript we will insert a dedicated derivation subsection that walks through the steps: (i) starting from the joint marginal p(data, partition), (ii) applying the posterior characterization to obtain the predictive integral, and (iii) contrasting the resulting time-dependent curve with the standard posterior mean of the cause-specific incidence function. We will reference the relevant equations and add a short remark clarifying why the curve constitutes a distinct predictive functional. revision: yes

Circularity Check

0 steps flagged

No circularity: prediction curve derived from posterior and marginals without reduction to fitted inputs

full rationale

The paper defines a class of hierarchical CRM priors for transition probabilities in a multi-state competing risks model, identifies the conditionally conjugate member, derives the joint marginal of data and latent partition, and characterizes the posterior. It then computes the predictive probability of event type as a function of time (the prediction curve) by leveraging these results. This is a standard Bayesian predictive derivation; the curve is not equivalent by construction to any fitted parameter or input quantity, nor does it rely on self-citation for its load-bearing steps. The conjugacy choice is an explicit modeling decision whose implications are separate from circularity. No quoted step reduces the claimed innovation to a renaming or tautology.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Central claim rests on standard properties of completely random measures and their hierarchical extensions, which are domain assumptions from Bayesian nonparametric literature; no explicit free parameters or new invented entities are stated in the abstract.

axioms (1)
  • domain assumption Hierarchical completely random measures can be used to define flexible nonparametric priors on transition probabilities in multi-state models
    Invoked to introduce the general class of priors for competing risks.

pith-pipeline@v0.9.0 · 5518 in / 1220 out tokens · 57730 ms · 2026-05-07T12:35:27.660117+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

76 extracted references · 1 canonical work pages

  1. [1]

    Andersen, P. K., S. Z. Abildstrom, and S. Rosthøj (2002). Competing Risks as a Multi--State Model . Statistical Methods in Medical Research\/ 11 , 203--215

    2002

  2. [2]

    Andersen, P. K., O. Borgan, R. D. Gill, and N. Keiding (1993). Statistical Models Based on Counting Processes . New York: Springer

    1993

  3. [3]

    Peluso, and P

    Arf\'e, A., S. Peluso, and P. Muliere (2019). Reinforced Urns and the Subdistribution Beta--Stacy Process for Competing Risks Analysis . Scandinavian Journal of Statistics\/ 46 , 706--734

    2019

  4. [4]

    Lijoi, L

    Barrios, E., A. Lijoi, L. E. Nieto-Barajas, and I. Pr\"unster (2013). Modeling with Normalized Random Measure Mixture Models . Statistical Science\/ 28 , 313--334

    2013

  5. [5]

    Latouche, A

    Beyersmann, J., A. Latouche, A. Buchholz, and M. Schumacher (2009). Simulating Competing Risks Data in Survival Analysis . Statistics in Medicine\/ 28 , 956--971

    2009

  6. [6]

    Blackwell, D. and J. B. MacQueen (1973). Ferguson Distributions via P\'olya Urn Schemes . Annals of Statistics\/ 1 , 353--355

    1973

  7. [7]

    Br\"uck, F. (2025). Generative Neural Networks for Characteristic Functions . Journal of Computational and Graphical Statistics\/ 34\/ (4), 1456--1465

    2025

  8. [8]

    Lijoi, P

    Camerlenghi, F., A. Lijoi, P. Orbanz, and I. Pr\"unster (2019). Distribution Theory for Hierarchical Processes . Annals of Statistics\/ 47 , 67--92

    2019

  9. [9]

    Lijoi, and I

    Camerlenghi, F., A. Lijoi, and I. Pr\"unster (2021). Survival Analysis via Hierarchically Dependent Mixture Hazards . Annals of Statistics\/ 49 , 863--884

    2021

  10. [10]

    Cassese, A., W. Zhu, M. Guindani, and M. Vannucci (2019). A Bayesian Nonparametric Spiked Process Prior for Dynamic Model Selection . Bayesian Analysis\/ 14 , 553--572

    2019

  11. [11]

    Catalano, M. and C. Del Sole (2025). Hierarchical Random Measures without Tables . arXiv:2505.02653

    work page arXiv 2025

  12. [12]

    Lijoi, and I

    Catalano, M., A. Lijoi, and I. Pr\"unster (2020). Approximation of Bayesian Models for Time--to--Event Data . Electronic Journal of Statistics\/ 14 , 3366--3395

    2020

  13. [13]

    Chen, C., V. Rao, W. Buntine, and Y. Whye Teh (2013). Dependent Normalized Random Measures . In Proceedings of Machine Learning Research , Volume 28, pp.\ 969--977

    2013

  14. [14]

    Chipman, H. A., E. I. George, and R. E. McCulloch (2010). BART: Bayesian Additive Regression Trees . Annals of Applied Statistics\/ 4 , 266--298

    2010

  15. [15]

    Christensen, J. and L. Ma (2020). A Bayesian Hierarchical Model for Related Densities by Using P\'olya Trees . Journal of the Royal Statistical Society: Series B\/ 82 , 127--153

    2020

  16. [16]

    Cook, R. J. and J. F. Lawless (2018). Multistate Models for the Analysis of Life History Data . Boca Raton: Chapman and Hall

    2018

  17. [17]

    Cox, D. R. (1959). The Analysis of Exponentially Distributed Life--Times with Two Types of Failures . Journal of the Royal Statistical Society: Series B\/ 21 , 411--421

    1959

  18. [18]

    Crowder, M. J. (1991). On the Identifiability Crisis in Competing Risks Analysis . Scandinavian Journal of Statistics\/ 18 , 223--233

    1991

  19. [19]

    Crowder, M. J. (2012). Multivariate Survival Analysis and Competing Risks . Chapman & Hall

    2012

  20. [20]

    Daley, D. J. and D. Vere-Jones (2007). An Introduction to the Theory of Point Processes: Volume II: General Theory and Structure . New York: Springer

    2007

  21. [21]

    Peccati, and I

    De Blasi, P., G. Peccati, and I. Pr\"unster (2009). Asymptotics for Posterior Hazards . Annals of Statistics\/ 37 , 1906--1945

    2009

  22. [22]

    Drzewiecki, K. T., C. Ladefoged, and H. E. Christensen (1980). Biopsy and Prognosis for Cutaneous Malignant Melanomas in Clinical Stage I . Scandinavian Journal of Plastic and Reconstructive Surgery\/ 14 , 141--144

    1980

  23. [23]

    Dykstra, R. L. and P. Laud (1981). A Bayesian Nonparametric Approach to Reliability . Annals of Statistics\/ 9 , 356--367

    1981

  24. [24]

    Epifani, I. and A. Lijoi (2010). Nonparametric Priors for Vectors of Survival Functions . Statistica Sinica\/ 20 , 1455--1484

    2010

  25. [25]

    Escobar, M. D. and M. West (1998). Computing Nonparametric Hierarchical Models . In Practical Nonparametric and Semiparametric Bayesian Statistics , pp.\ 1--22. New York: Springer

    1998

  26. [26]

    Ferguson, T. S. and M. J. Klass (1972). A Representation of Independent Increment Processes without Gaussian Components . Annals of Mathematical Statistics\/ 43 , 1634--1643

    1972

  27. [27]

    Fine, J. P. and R. J. Gray (1999). A Proportional Hazards Model for the Subdistribution of a Competing Risk . Journal of the American Statistical Association\/ 94 , 496--509

    1999

  28. [28]

    Geskus, R. B. (2015). Data Analysis with Competing Risks and Intermediate States . Chapman & Hall

    2015

  29. [29]

    Geskus, R. B. (2024). Competing Risks: Concepts, Methods, and Software . Annual Review of Statistics and Its Application\/ 11 , 227--254

    2024

  30. [30]

    Ghosal, S. and A. van der Vaart (2017). Fundamentals of Nonparametric Bayesian Inference . Cambridge: Cambridge University Press

    2017

  31. [31]

    Griffin, J. E. and F. Leisen (2017). Compound Random Measures and Their Use in Bayesian Non--Parametrics . Journal of the Royal Statistical Society: Series B\/ 79 , 525--545

    2017

  32. [32]

    Hjort, N. L. (1990). Nonparametric Bayes Estimators Based on Beta Processes in Models for Life History Data . Annals of Statistics\/ 18 , 1259--1294

    1990

  33. [33]

    Ishwaran, H. and L. F. James (2004). Computational Methods for Multiplicative Intensity Models Using Weighted Gamma Processes: Proportional Hazards, Marked Point Processes, and Panel Count Data . Journal of the American Statistical Association\/ 99 , 175--190

    2004

  34. [34]

    James, L. F. (2005). Bayesian Poisson Process Partition Calculus with an Application to Bayesian L\'evy Moving Averages . Annals of Statistics\/ 33 , 1771--1799

    2005

  35. [35]

    James, L. F., A. Lijoi, and I. Pr\"unster (2009). Posterior Analysis for Normalized Random Measures with Independent Increments . Scandinavian Journal of Statistics\/ 36 , 76--97

    2009

  36. [36]

    Jara, A. and T. E. Hanson (2011). A Class of Mixtures of Dependent Tail--Free Processes . Biometrika\/ 98\/ (3), 553--566

    2011

  37. [37]

    Lesaffre, M

    Jara, A., E. Lesaffre, M. De Iorio, and F. Quintana (2010). Bayesian Semiparametric Inference for Multivariate Doubly--Interval--Censored Data . Annals of Applied Statistics\/ 4 , 2126--2149

    2010

  38. [38]

    Kalbfleisch, J. D. and R. L. Prentice (2002). The Statistical Analysis of Failure Time Data . Wiley

    2002

  39. [39]

    Kingman, J. F. (1967). Completely Random Measures . Pacific Journal of Mathematics\/ 21 , 59--78

    1967

  40. [40]

    Kingman, J. F. (1993). Poisson Processes . Oxford: Clarendon Press

    1993

  41. [41]

    Lawless, J. F. (2003). Statistical Models and Methods for Lifetime Data . Wiley

    2003

  42. [42]

    Lijoi, A. and B. Nipoti (2014). A Class of Hazard Rate Mixtures for Combining Survival Data from Different Experiments . Journal of the American Statistical Association\/ 109 , 802--814

    2014

  43. [43]

    Nipoti, and I

    Lijoi, A., B. Nipoti, and I. Pr\"unster (2014). Bayesian Inference with Dependent Normalized Completely Random Measures . Bernoulli\/ 20 , 1260--1291

    2014

  44. [44]

    Lijoi, A. and I. Pr\"unster (2010). Models beyond the Dirichlet process . In Bayesian Nonparametrics , pp.\ 80--136. Cambridge University Press

    2010

  45. [45]

    Pr\"unster, and T

    Lijoi, A., I. Pr\"unster, and T. Rigon (2020). Sampling Hierarchies of Discrete Random Structures . Statistics and Computing\/ 30 , 1591--1607

    2020

  46. [46]

    and C.-S

    Lo, A. and C.-S. Weng (1989). On a Class of Bayesian Nonparametric Estimates: II. Hazard Rate Estimates . Annals of the Institute of Statistical Mathematics\/ 41 , 227--245

    1989

  47. [47]

    MacEachern, S. N. (1998). Computational Methods for Mixture of Dirichlet Process Models . In Practical Nonparametric and Semiparametric Bayesian Statistics , pp.\ 23--43. New York: Springer

    1998

  48. [48]

    M\"uller, P. and F. Quintana (2004). Nonparametric Bayesian Data Analysis . Statistical Science\/ 19 , 95--110

    2004

  49. [49]

    Quintana, A

    M\"uller, P., F. Quintana, A. Jara, and T. E. Hanson (2015). Bayesian Nonparametric Data Analysis . Cham: Springer

    2015

  50. [50]

    Quintana, and G

    M\"uller, P., F. Quintana, and G. L. Rosner (2004). A Method for Combining Inference across Related Nonparametric Bayesian Models . Journal of the Royal Statistical Society: Series B\/ 66 , 735--749

    2004

  51. [51]

    Nieto-Barajas, L. E. and S. G. Walker (2004). Bayesian Nonparametric Survival Analysis via L\'evy Driven Markov Processes . Statistica Sinica\/ 14\/ (4), 1127--1146

    2004

  52. [52]

    Jara, and M

    Nipoti, B., A. Jara, and M. Guindani (2018). A Bayesian Semiparametric Partially PH Model for Clustered Time--to--Event Data . Scandinavian Journal of Statistics\/ 45 , 1016--1035

    2018

  53. [53]

    Oganisian, A., K. D. Getz, T. A. Alonzo, R. Aplenc, and J. A. Roy (2024). Bayesian Semiparametric Model for Sequential Treatment Decisions with Informative Timing . Biostatistics\/ 25 , 947--961

    2024

  54. [54]

    Passeggeri, R. (2023). On Quasi--Infinitely Divisible Random Measures . Bayesian Analysis\/ 18 , 253--286

    2023

  55. [55]

    Passeggeri, R. (2025). Random Signed Measures . Bayesian Analysis\/ (forthcoming)

    2025

  56. [56]

    Peccati, G. and I. Pr\"unster (2008). Linear and Quadratic Functionals of Random Hazard Rates: An Asymptotic Analysis . Annals of Applied Probability\/ 18 , 1910--1943

    2008

  57. [57]

    Pitman, J. (1996). Some Developments of the Blackwell--MacQueen Urn Scheme . In Statistics, Probability and Game Theory , pp.\ 245--267. IMS

    1996

  58. [58]

    Fiocco, and R

    Putter, H., M. Fiocco, and R. B. Geskus (2007). Tutorial in Biostatistics: Competing Risks and Multi--State Models . Statistics in Medicine\/ 26 , 2389--2430

    2007

  59. [59]

    Quintana, F. A., P. M\"uller, A. Jara, and S. N. MacEachern (2022). The Dependent Dirichlet Process and Related Models . Statistical Science\/ 37 , 24--41

    2022

  60. [60]

    Lijoi, and I

    Regazzini, E., A. Lijoi, and I. Pr\"unster (2003). Distributional Results for Means of Normalized Random Measures with Independent Increments . Annals of Statistics\/ 31 , 560--585

    2003

  61. [61]

    Riva-Palacio, A. and F. Leisen (2018). Bayesian Nonparametric Estimation of Survival Functions with Multiple--Samples Information . Electronic Journal of Statistics\/ 12 , 1330--1357

    2018

  62. [62]

    Riva-Palacio, A. and F. Leisen (2021). Compound Vectors of Subordinators and Their Associated Positive L\'evy Copulas . Journal of Multivariate Analysis\/ 183 , 104728

    2021

  63. [63]

    Leisen, and J

    Riva-Palacio, A., F. Leisen, and J. E. Griffin (2022). Survival Regression Models with Dependent Bayesian Nonparametric Priors . Journal of the American Statistical Association\/ 117 , 1530--1539

    2022

  64. [64]

    Sato, K.-I. (1999). L \'e vy Processes and Infinitely Divisible Distributions . Cambridge: Cambridge University Press

    1999

  65. [65]

    Buchholz, and A.-K

    Schmitt, S., A. Buchholz, and A.-K. Ozga (2023). Systematic Comparison of Approaches to Analyze Clustered Competing Risks Data . BMC Medical Research Methodology\/ 23 , 86

    2023

  66. [66]

    Scoggins, C. R., M. I. Ross, D. S. Reintgen, D. Noyes, J. S. Goydos, P. D. Beitsch, M. M. Urist, S. Ariyan, J. J. Sussman, M. J. Edwards, A. B. Chagpar, R. C. G. Martin, A. J. Stromberg, L. Hagendoorn, and K. M. McMasters (2006). Gender--Related Differences in Outcome for Melanoma Patients . Annals of Surgery\/ 243 , 693--700

    2006

  67. [67]

    Soriano, J. and L. Ma (2019). Mixture Modeling on Related Samples by --Stick Breaking and Kernel Perturbation . Bayesian Analysis\/ 14 , 161--180

    2019

  68. [68]

    Sparapani, R., B. R. Logan, R. E. McCulloch, and P. W. Laud (2020). Nonparametric Competing Risks Analysis Using Bayesian Additive Regression Trees . Statistical Methods in Medical Research\/ 29 , 57--77

    2020

  69. [69]

    Sparapani, R. A., B. R. Logan, R. E. McCulloch, and P. W. Laud (2016). Nonparametric Survival Analysis Using Bayesian Additive Regression Trees (BART) . Statistics in Medicine\/ 35 , 2741--2753

    2016

  70. [70]

    Teh, Y. W., M. I. Jordan, M. J. Beal, and D. M. Blei (2006). Hierarchical Dirichlet Processes . Journal of the American Statistical Association\/ 101 , 1566--1581

    2006

  71. [71]

    Tsiatis, A. (1975). A Nonidentifiability Aspect of the Problem of Competing Risks . Proceedings of the National Academy of Sciences\/ 72 , 20--22

    1975

  72. [72]

    Walker, S. and P. Muliere (1997). Beta--Stacy Processes and a Generalization of the P\'olya--Urn Scheme . Annals of Statistics\/ 25 , 1762--1780

    1997

  73. [73]

    Wolpert, R. L. and K. Ickstadt (1998). Simulation of L\'evy Random Fields . In Practical Nonparametric and Semiparametric Bayesian Statistics , pp.\ 227--242. Springer New York

    1998

  74. [74]

    M\"uller, A

    Xu, Y., P. M\"uller, A. S. Wahed, and P. F. Thall (2016). Bayesian Nonparametric Estimation for Dynamic Treatment Regimes with Sequential Transition Times . Journal of the American Statistical Association\/ 111 , 921--950

    2016

  75. [75]

    Scharfstein, P

    Xu, Y., D. Scharfstein, P. M\"uller, and M. Daniels (2020). A Bayesian Nonparametric Approach for Evaluating the Causal Effect of Treatment in Randomized Trials with Semi-Competing Risks . Biostatistics\/ 23 , 34--49

    2020

  76. [76]

    Zhou, B., J. Fine, A. Latouche, and M. Labopin (2012). Competing Risks Regression for Clustered Data . Biostatistics\/ 13 , 371--383

    2012