betaFeedback

WorksMathematics

Proximal R2 Priors for Structured Variance Allocation

Abstract

Scientific domain knowledge in regression often concerns structural relationships among model terms, which are typically encoded as parameter constraints. Since such relationships rarely hold exactly in practice, prior distributions should encourage them while at the same time retaining support for scientifically plausible departures. Accounting for these relationships becomes more pressing in modern regression, where high-dimensionality is usually present and regularisation is essential. R2R^2 decomposition priors provide an interpretable foundation for regularisation by placing a prior on total explained variance and distributing the implied signal across model terms. We introduce the Proximal R2 prior, which uses proximal relaxations to encode these relationships as soft constraints on the variance allocation. Our formulation yields differentiable posteriors compatible with standard gradient based samplers. Theoretically, we prove that our prior preserves the characteristic marginal properties of global-local shrinkage priors. Moreover, we establish posterior consistency and contraction rates in growing dimensions, showing that the inclusion of structural relationships and proximal methods do not alter the shrinkage properties. Notably, these results demonstrate that the structural constraints in the model do not hinder posterior recovery. Simulation studies and real-data applications to interaction and generalised additive models demonstrate that our prior improves parameter recovery, out-of-sample prediction, and sampling efficiency over existing constrained and unconstrained shrinkage priors, even under structural misspecification.

Introduction

Modern regression models frequently contain terms linked by scientifically meaningful relationships. In an interaction model, for instance, an interaction is often considered more plausible when the main effects are active, a principle known as heredity [22, 34]. In a generalised additive model, a parsimonious linear effect might be preferred unless the data provide sufficient evidence for a non-linear departure [41, 51]. Likewise, domain knowledge often suggests that certain treatment exposures or mechanisms should receive greater prior signal than others [37, 74]. Such structural beliefs govern the relative relevance of model terms, yet they rarely hold exactly in practice. A useful prior should encourage the proposed structure without preventing the data from supporting scientifically credible departures from it [72].

Incorporating such structural guidance is especially critical in high-dimensional regression problems, where the number of covariates is large relative to the sample size [81]. Bayesian inference responds to the complexity of these problems by designing prior distributions that encode structure [20, 19]. Global-local shrinkage priors provide a natural starting point to encourage proposed structures since they combine model-wide regularization with term-specific adaptation [18, 9, 13, 66, 11] (Polson and Scott 2012). An array of shrinkage priors encode relationships among the local scales, facilitating regularization of one model term dependent on its relationship to others [88, 35, 6]. R2 type priors allow for interpretable model-level prior specification by separating priors beliefs about total explained variance, R2R^2, from signal allocation to model terms [89]. This provides a principled route for incorporating structure into the model while still allowing domain knowledge to be easily elicited. Originally designed for ordinary linear regression, the framework has since been extended to multilevel [2, 84], time-series [45], ordinal [84], spatial [85] and grouped [6, 87] settings. Despite these advances, a general framework for encoding flexible parameter constraints and relational preferences directly into the signal allocation mechanism remains lacking.

To this end, we develop an R2 prior which embeds prespecified structural relationships among model terms and parameter constraints directly as preferences using proximal methods [69]. In view of this, we name our prior distribution Proximal R2, which we denote by Prox-R2. Our prior retains the model-level interpretation of the R2 prior similar to Zhang et al. [89], while treating relationships as guidance, rather than properties that must hold exactly. Furthermore, the uncertainty related to constraints and relationships is expressed via proximal relaxations of the corresponding hard-constrained prior, i.e., one that would enforce the constraints strictly ([69, 23, 70, 90]). Our approach results in a family of softly constrained posteriors which approach the hard-constrained formulation as the amount of smoothing decreases; while still retaining support for scientifically plausible departures from the preferred relationships. Moreover, our models can be sampled efficiently with off-the-shelf state-of-the-art gradient-based samplers ([55, 42]); a consequence of using proximal methods.

The combination of dependence and parametric constraints raises the question whether the additional model structure preserves typical shrinkage prior behaviour, namely, strong shrinkage of noise and signal detection ([67, 59, 78]). While explicit dependence among proportions of variance has been considered before ([6]), their interaction with explicit structural relationships has not been studied. In particular, we focus on problems whose structural relationships of interest can be expressed as sets of linear constraints (although our framework is sufficiently flexible to account for general forms). These constraints induce a non-standard joint distribution on the simplex, whose marginal allocation densities are generally unavailable in closed form, making their implications for shrinkage difficult to anticipate. We show that the proposed prior preserves the desired marginal shrinkage behaviour. Our analysis characterises how the global scale and the distribution of allocation shares interact to jointly determine the marginal coefficient density near zero and in the tails.

We also establish posterior consistency and contraction for Gaussian regression with growing dimension, importantly, without requiring the true coefficients to obey the favoured structural constraints. To the best of our knowledge, these represent the first theoretical results for shrinkage behaviour and posterior consistency and contraction for continuous shrinkage priors coupled with proximal methods. The theoretical contribution also extends to the unconstrained logistic-normal R2 prior ([6]), for which the marginal origin, tail asymptotics, and posterior contraction rates had not previously been established.

We demonstrate the practical utility and scalability of the Prox-R2 prior through simulations and real-data applications involving interaction and additive models. Across the considered regimes, Prox-R2 consistently outperforms established shrinkage priors, regardless of whether they take constraints into account or not in their formulations. The performance gains are most pronounced under moderate-to-high signal-to-noise ratios, especially for identifying non-zero coefficients. Our method remains useful even when the constraints hold only approximately. Our results illustrate how structural knowledge, even when uncertain, can guide high-dimensional regularisation while retaining interpretable control in complex high-dimensional settings.

The paper is organized as follows. Section 2 introduces the proposed prior while Section 3 describes the computational approach. The theoretical results are presented in Section 4 with Section 5 presenting the simulation and real-data results. Finally, Section 6 discusses the findings and outlines directions for future research.

Methodology

Model

Given responses yiy_i, covariates xi′=(xi1,…,xip)\mathbf{x}_i^{\prime}=(x_{i1},\ldots,x_{ip}) and zi=h(xi)∈Rd\mathbf{z}_i=h(\mathbf{x}_i)\in\mathbb{R}^d, which collects model terms constructed from the original covariates, consider the additive model

yi=zi′β+εi,(1)y_i=\mathbf{z}_i^{\prime}\boldsymbol{\beta}+\varepsilon_i, \tag*{(1)}

where εi∼iid.Normal⁡(0,σ2)\varepsilon_i\overset{\mathrm{iid.}}{\sim}\operatorname{Normal}(0,\sigma^2) for i=1,…,ni=1,\ldots,n. In ordinary linear regression, h(xi)=xih(\mathbf{x}_i)=\mathbf{x}_i and d=pd=p while in models that contain added structure, h(⋅)h(\cdot) may include interactions, spline basis functions, polynomial terms, or other transformations (Griffin and Brown [34]).

R2 priors

We adopt the global–local shrinkage framework ([78] (Polson and Scott 2012)), in which the prior for each regression coefficient is represented as a scale mixture of normals ([8, 82]). Specifically, conditional on the coefficient-specific mixing scales, the regression coefficients are independent and Gaussian:

βj∣σ2,ψj2∼Normal⁡(0,σ2ψj2),j=1,…,d,(2)\beta_j\mid\sigma^2,\psi_j^2\sim\operatorname{Normal}(0,\sigma^2\psi_j^2),\qquad j=1,\ldots,d, \tag*{(2)}

where ψj2\psi_j^2 is the coefficient-specific prior scale, which we will decompose into global signal and term-specific allocation parameters.

Following Zhang et al. [89], we construct a prior for a model-level coefficient-of-determination R2R^2 which induces a prior distribution on each individual ψj2\psi_j^2. Treating z\mathbf{z} as random and independent of β\boldsymbol{\beta}, we define R2R^2 under the joint distribution induced by the coefficient prior and the covariate distribution as

R2=corr⁡2(y,z′β)=Var⁡(z′β)Var⁡(z′β)+σ2.(3)R^2=\operatorname{corr}^2(y,\mathbf{z}^{\prime}\boldsymbol{\beta})=\frac{\operatorname{Var}(\mathbf{z}^{\prime}\boldsymbol{\beta})}{\operatorname{Var}(\mathbf{z}^{\prime}\boldsymbol{\beta})+\sigma^2}. \tag*{(3)}

Using the law of total variance, we obtain

Var⁡(z′β)=Ez{Var⁡(z′β∣z)}+Var⁡z{E(z′β∣z)}(4)\operatorname{Var}(\mathbf{z}^{\prime}\boldsymbol{\beta})=\mathbb{E}_{\mathbf{z}}\{\operatorname{Var}(\mathbf{z}^{\prime}\boldsymbol{\beta}\mid\mathbf{z})\}+\operatorname{Var}_{\mathbf{z}}\{\mathbb{E}(\mathbf{z}^{\prime}\boldsymbol{\beta}\mid\mathbf{z})\} \tag*{(4)}

where the second term is zero and Var⁡(z′β∣z)=σ2∑j=1dzj2ψj2\operatorname{Var}(\mathbf{z}^{\prime}\boldsymbol{\beta}\mid\mathbf{z})=\sigma^2\sum_{j=1}^{d}z_j^2\psi_j^2 such that

Var⁡(z′β)=σ2∑j=1dψj2Ez(zj2).(5)\operatorname{Var}(\mathbf{z}^{\prime}\boldsymbol{\beta})=\sigma^2\sum_{j=1}^{d}\psi_j^2\mathbb{E}_{\mathbf{z}}(z_j^2). \tag*{(5)}

We assume that Ez(zj2)<∞\mathbb{E}_{\mathbf{z}}(z_j^2)<\infty for all jj. To separate the allocation of prior signal from the scale of the associated model term, we define ρj=Ez(zj2)\rho_j=\mathbb{E}_{\mathbf{z}}(z_j^2) for j=1,…,dj=1,\ldots,d (Aguilar and Bürkner [2]), and parametrise the coefficient-specific variances as

ψj2=ϕjω2ρj,j=1,…,d,(6)\psi_j^2=\frac{\phi_j\omega^2}{\rho_j},\qquad j=1,\ldots,d, \tag*{(6)}

where ϕ=(ϕ1,…,ϕd)′\phi= (\phi_1,\ldots,\phi_d)' lies on the simplex, i.e ϕj≥0\phi_j \ge0 and ∑j=1dϕj=1\sum_{j=1}^{d}\phi_j = 1. Then ∑j=1dψj2Ez(zj2)=ω2\sum_{j=1}^{d}\psi_j^2\mathbb{E}_z(z_j^2) = \omega^2 such that

R2=σ2ω2σ2ω2+σ2=ω2ω2+1.(7)R^2 = \frac{\sigma^2\omega^2}{\sigma^2\omega^2+\sigma^2} = \frac{\omega^2}{\omega^2+1}. \tag*{(7)}

Therefore, ω2\omega^2 controls the total prior signal relative to the residual variance, while ϕj\phi_j is the variance allocation share assigned to the jjth model term. As in previous work on R2 priors, we let R2R^2 follow a Beta distribution, which induces a beta-prime prior distribution [43] on ω2\omega^2 i.e., R2∼Beta⁡(a,b)R^2 \sim\operatorname{Beta}(a,b) implies ω2∼BetaPrime⁡(a,b)\omega^2 \sim\operatorname{BetaPrime}(a,b).

Logistic-normal allocation

The R2 prior framework requires that ϕ\phi lie on the simplex so that the total prior signal ω2\omega^2 is apportioned to the different model terms [89]. Generally speaking, if (T1,…,Td)(T_1,\ldots,T_d) is a vector of positive random variables, then we can define

ϕj=Tj∑r=1dTr,j=1,…,d.(8)\phi_j = \frac{T_j}{\sum_{r=1}^{d}T_r}, \qquad j=1,\ldots,d. \tag*{(8)}

For example, taking TjT_j's to be independent gamma random variables with a common rate parameter induces a Dirichlet distribution on ϕ\phi (e.g., [46, 48, 89, 6]). Alternatively, we can consider a latent Gaussian vector η∼Normal⁡(μη,Ση)\eta\sim\operatorname{Normal}(\mu_\eta,\Sigma_\eta) and map it to the simplex through the softmax operator, thus setting Tj=exp⁡(ηj)T_j = \exp(\eta_j). The resulting distribution of ϕ\phi is a logistic-normal ([7, 6, 87]). Through the covariance matrix Ση\Sigma_\eta, the logistic-normal distribution can accommodate a richer range of dependence structures among the components of ϕ\phi than the Dirichlet distribution, whose covariance structure is fully determined by its mean vector and concentration vector.

Although constraints could be imposed directly on ϕ\phi, doing so entangles the scientific restrictions with the geometry of the simplex. In particular, the constraints must be compatible with both positivity and the unit-sum condition. Restricting any one component necessarily changes the feasible ranges of the others. Such restrictions can therefore alter the marginal behavior and dependence structure of the entire composition in ways that are difficult to anticipate. A construction based on positive variables (T1,…,Td)(T_1,\ldots,T_d) does not fully resolve this issue. Even when the joint distribution is specified flexibly, normalization can obscure the implications for the induced distribution of ϕ\phi [46].

For these reasons, we adopt the logistic-normal construction and impose constraints on the latent Gaussian parameter η\eta rather than directly on ϕ\phi. The softmax transformation automatically enforces positivity and the unit-sum condition, thereby separating the compositional normalization from the scientific restrictions of interest. Moreover, the identity log⁡(ϕj/ϕk)=ηj−ηk\log(\phi_j/\phi_k) = \eta_j-\eta_k provides direct control over relative allocations. Consequently, ordering constraints, bounds on relative magnitudes, and more general multiplicative relationships among the components of ϕ\phi can often be represented as linear constraints on contrasts of η\eta ([6]). Finally, imposing restrictions on η\eta directly, does not alter the distribution of R2R^2 unlike working directly with the regression coefficients β\beta.

Motivating examples

We now describe several modelling situations in which constraints on the latent allocation parameters naturally arise.

Domain expertise. One of the simplest examples comes from domain information where practitioners believe that the prior variance assigned to one term should be bounded by that assigned to another. In the latent scale, this means

ηj≤ηk.(9)\eta_j \le\eta_k. \tag*{(9)}

Since the softmax denominator is common to all coordinates, this is equivalent to ϕj≤ϕk\phi_j \le\phi_k. More generally, if the prior allocation of term jj is expected to be at most c>0c > 0 times the allocation of term kk, then we may impose

ηj−ηk≤log⁡c,(10)\eta_j - \eta_k \le\log c, \tag*{(10)}

which is equivalent to ϕj≤cϕk\phi_j \le c\phi_k.

Heredity in interaction models. Interaction models are another common source of variance constraints ([22, 34]). Consider a regression model containing all main effects and second-order interactions,

yi=∑j=1pxijβj+∑j<kxijxikβjk+εi.(11)y_i = \sum_{j=1}^{p} x_{ij}\beta_j + \sum_{j<k} x_{ij}x_{ik}\beta_{jk} + \varepsilon_i. \tag*{(11)}

Letting zi=(xi1,…,xip,xi1xi2,xi1xi3,…,xi,p−1xip)′z_i = (x_{i1}, \ldots, x_{ip}, x_{i1}x_{i2}, x_{i1}x_{i3}, \ldots, x_{i,p-1}x_{ip})^\prime and the vector of coefficients β=(β1,…,βp,β12,β13,…,βp−1 p)\beta= (\beta_1, \ldots, \beta_p, \beta_{12}, \beta_{13}, \ldots, \beta_{p-1\,p}) places this model in the general form of (1), with d=p+p(p−1)/2d = p + p(p - 1)/2 model terms.

Heredity expresses the prior belief that an interaction should receive considerable signal only when the corresponding main effects do ([39]). On the latent scale, we encode this belief by requiring the interaction parameter to be no greater than the average of the corresponding main-effect parameters:

ηjk≤12(ηj+ηk),j<k,(12)\eta_{jk} \le\frac{1}{2}(\eta_j + \eta_k), \qquad j < k, \tag*{(12)}

References

  1. [1]Abril-Pla, O., Andreani, V., Carroll, C., Dong, L., Fonnesbeck, C. J., Kochurov, M., Kumar, R., Lao, J., Luhmann, C. C., Martin, O. A., Osthege, M., Vieira, R., Wiecki, T. and Zinkov, R. (2023) PyMC: A Modern and Comprehensive Probabilistic Programming Framework in Python. PeerJ Computer Science, 9.DOI
  2. [2]Aguilar, J. E. and Bürkner, P.-C. (2023) Intuitive joint priors for Bayesian linear multilevel models: The R2D2M2 prior. Electronic Journal of Statistics, 17, 1711 – 1767. URL: https://doi.org/10.1214/23-EJS2136.
  3. [5]Aguilar, J. E., Klippert, D., Balci, Y. and Marx, A. (2026) Dirichlet-tree allocation priors for bayesian causal forests. URL: https://arxiv.org/abs/2610.26728.
  4. [6]Aguilar, J. E., Kohns, D., Vehtari, A. and Bürkner, P.-C. (2025) R2 priors for Grouped Variance Decomposition in High-dimensional Regression. arXiv preprint arXiv:2507.11833.arxiv.org/abs/2507.11833
  5. [7]Aitchison, J. and Shen, S. M. (1980) Logistic-Normal Distributions: Some Properties and Uses. Biometrika, 67, 261–272. URL: http://www.jstor.org/stable/2335470.DOI
  6. [8]Andrews, D. F. and Mallows, C. L. (1974) Scale mixtures of normal distributions. Journal of the Royal Statistical Society: Series B (Methodological), 36, 99–102.DOI
  7. [9]Armagan, A., Clyde, M. and Dunson, D. B. (2011) Generalized beta mixtures of Gaussians. In Advances in neural information processing systems, 523–531.arxiv.org/abs/1107.4976
  8. [10]Armagan, A., Dunson, D. B., Lee, J., Bajwa, W. U. and Strawn, N. (2013) Posterior consistency in linear models under shrinkage priors. Biometrika, 100, 1011–1018. URL: https://doi.org/10.1093/biomet/ast028.
  9. [11]Bai, R. and Ghosh, M. (2021) On the Beta Prime Prior for Scale Parameters in High-Dimensional Bayesian Regression Models. Statistica Sinica. URL: http://arxiv.org/abs/1807.06539.
  10. [12]Barndorff-Nielsen, O., Kent, J. and Sørensen, M. (1982) Normal variance-mean mixtures and \z\ distributions. International Statistical Review, 50, 145–159.
  11. [13]Bhattacharya, A., Pati, D., Pillai, N. S. and Dunson, D. B. (2015) Dirichlet–Laplace Priors for Optimal Shrinkage. Journal of the American Statistical Association, 110, 1479–1490. URL: https://doi.org/10.1080/01621459.2014.960967.
  12. [14]Bockting, F., Radev, S. T. and Bürkner, P.-C. (2024) Simulation-based prior knowledge elicitation for parametric Bayesian models. Scientific Reports, 14, 17330. URL: https://www.nature.com/articles/s41598-024-68090-7.
  13. [15]Boss, J., Datta, J., Wang, X., Park, S. K., Kang, J. and Mukherjee, B. (2023) Group Inverse-Gamma Gamma Shrinkage for Sparse Linear Models with Block-Correlated Regressors. Bayesian Analysis, 1 – 30. URL: https://doi.org/10.1214/23-BA1371.
  14. [16]Botev, Z. I. (2017) The Normal Law Under Linear Restrictions: Simulation and Estimation via Minimax Tilting. Journal of the Royal Statistical Society Series B: Statistical Methodology, 79, 125–148. URL: http://arxiv.org/abs/1603.04166. ArXiv:1603.04166 [stat].
  15. [17]Breiman, L. and Friedman, J. H. (1985) Estimating optimal transformations for multiple regression and correlation. Journal of the American Statistical Association, 80, 580–598.
  16. [18]Carvalho, C. M., Polson, N. G. and Scott, J. G. (2010) The horseshoe estimator for sparse signals. Biometrika, 97, 465–480. URL: http://www.jstor.org/stable/25734098.
  17. [19]Castillo, I., Schmidt-Hieber, J. and Vaart, A. W. v. d. (2015) Bayesian Linear Regression with Sparse Priors. The Annals of Statistics, 43, 1986–2018. URL: https://www.jstor.org/stable/43818568.DOI
  18. [20]Castillo, I. and Vaart, A. W. v. d. (2012) Needles and Straw in a Haystack: Posterior concentration for possibly sparse sequences. The Annals of Statistics, 40, 2069 – 2101. URL: https://doi.org/10.1214/12-AOS1029.
  19. [21]Chaari, L., Tourneret, J.-Y., Chaux, C. and Batatia, H. (2015) A hamiltonian monte carlo method for non-smooth energy sampling. URL: https://arxiv.org/abs/1401.3988.
  20. [22]Chipman, H., Hamada, M. and Wu, C. (1997) A Bayesian variable-selection approach for analyzing designed experiments with complex aliasing. Technometrics, 39, 372–381.DOI
  21. [23]Duan, L. L., Young, A. L., Nishimura, A. and Dunson, D. B. (2020) Bayesian constraint relaxation. Biometrika, 107, 191–204. URL: https://academic.oup.com/biomet/article/107/1/191/5686747.DOI
  22. [24]Durmus, A., Moulines, E. and Pereyra, M. (2016) Efficient bayesian computation by proximal markov chain monte carlo: when langevin meets moreau. URL: https://arxiv.org/abs/1612.07471.
  23. [25]Dykstra, R. L. (1983) An Algorithm for Restricted Least Squares Regression. Journal of the American Statistical Association, 78, 837–842. URL: http://www.jstor.org/stable/2288193.DOI
  24. [26]Echambadi, R. and Hess, J. D. (2007) Mean-centering does not alleviate collinearity problems in moderated multiple regression models. Marketing Science, 26, 438–445.DOI
  25. [27]Efron, B., Hastie, T., Johnstone, I. and Tibshirani, R. (2004) Least angle regression. The Annals of Statistics, 32, 407–499.
  26. [29]Gelman, A. (2008) Scaling regression inputs by dividing by two standard deviations. Statistics in Medicine, 27, 2865–2873.pubmed.ncbi.nlm.nih.gov/17960576
  27. [30]Gelman, A., Vehtari, A., Simpson, D., Margossian, C. C., Carpenter, B., Yao, Y., Kennedy, L., Gabry, J., Bürkner, P.-C. and Modrák, M. (2020) Bayesian Workflow. URL: https://arxiv.org/abs/2011.01808.
  28. [31]Ghosal, S. and van der Vaart, A. (2007) Convergence rates of posterior distributions for non-i.i.d. observations. The Annals of Statistics, 35, 192–223.
  29. [32]Ghosh, J. K. (2003) Bayesian Nonparametrics. Springer Series in Statistics Ser. New York, NY: Springer New York.
  30. [33]Ghosh, P. and Chakrabarti, A. (2017) Asymptotic Optimality of One-Group Shrinkage Priors in Sparse High-dimensional Problems. Bayesian Analysis, 12, 1133 – 1161. URL: https://doi.org/10.1214/16-BA1029.
  31. [34]Griffin, J. and Brown, P. (2017) Hierarchical Shrinkage Priors for Regression Models. Bayesian Analysis, 12, 135 – 159. URL: https://doi.org/10.1214/15-BA990.
  32. [35]Griffin, M. and Hoff, P. D. (2024) Structured Shrinkage Priors. Journal of Computational and Graphical Statistics, 33, 1–14. URL: https://www.tandfonline.com/doi/citedby/10.1080/10618600.2023.2233577.arxiv.org/abs/1902.05106
  33. [36]Guo, M., Huang, B., Yang, L. and Jiang, G. (2024) Above ground level estimation for radar altimetry using proximal hamiltonian monte carlo. IEEE Journal on Miniaturization for Air and Space Systems, 5, 128–137.DOI
  34. [37]Hahn, P. R. and Carvalho, C. M. (2015) Decoupling Shrinkage and Selection in Bayesian Linear Models: A Posterior Summary Perspective. Journal of the American Statistical Association, 110, 435–448. URL: https://doi.org/10.1080/01621459.2014.993077.
  35. [38]Hahn, P. R., Murray, J. S. and Carvalho, C. (2019) Bayesian regression tree models for causal inference: regularization, confounding, and heterogeneous effects. URL: https://arxiv.org/abs/1706.09523.
  36. [39]Haris, A., Witten, D. and Simon, N. (2016) Convex modeling of interactions with strong heredity. Journal of Computational and Graphical Statistics, 25, 981–1004. URL: http://dx.doi.org/10.1080/10618600.2015.1067217.
  37. [40]Harrison, D. and Rubinfield, D. L. (1978) Hedonic housing prices and the demand for clean air. Journal of Environmental Economics and Management, 5, 81–102.DOI
  38. [41]Hastie, T. and Tibshirani, R. (1986) Generalized Additive Models. Statistical Science, 1, 297–310. URL: http://www.jstor.org/stable/2245459.
  39. [42]Hoffman, M. D. and Gelman, A. (2014) The No-U-Turn Sampler: Adaptively Setting Path Lengths in Hamiltonian Monte Carlo. Journal of Machine Learning Research, 15, 1593–1623.DOI
  40. [43]Johnson, N. L., Kotz, S. and Balakrishnan, N. (1994) Continuous univariate distributions, volume 1, vol. 1. John wiley & sons.DOI
  41. [44]Joseph, V. R. (2006) A bayesian approach to the design and analysis of fractionated experiments. Technometrics, 48, 219–229. URL: http://www.jstor.org/stable/25471158.
  42. [45]Kohns, D., Kallioinen, N., McLatchie, Y. and Vehtari, A. (2026) The ARR2 prior: flexible predictive prior definition for Bayesian auto-regressions. Bayesian Analysis, 21, 1131–1162.
  43. [46]Kruijer, W., Rousseau, J. and Vaart, A. W. v. d. (2010) Adaptive Bayesian Density Estimation with Location-Scale Mixtures. Electronic Journal of Statistics, 4.
  44. [47]Lehmann, E. L. and Casella, G. (1998) Theory of Point Estimation. New York: Springer, 2 edn.
  45. [48]Lin, J. (2016) On The Dirichlet Distribution.
  46. [49]Linero, A. R. (2024) In Nonparametric and High-Dimensional Models, Bayesian Ignorability is an Informative Prior. Journal of the American Statistical Association, 119, 2785–2798.
  47. [50]Manderson, A. A. and Goudie, R. J. B. (2026) Translating predictive distributions into informative priors. Journal of the American Statistical Association, 0, 1–18. URL: https://doi.org/10.1080/01621459.2026.2614034. _eprint: https://doi.org/10.1080/01621459.2026.2614034.
  48. [51]Marra, G. and Wood, S. N. (2011) Practical variable selection for generalized additive models. Computational Statistics & Data Analysis, 55, 2372–2387.
  49. [52]McLatchie, Y., Rögnvaldsson, S., Weber, F. and Vehtari, A. (2023) Robust and efficient projection predictive inference.
  50. [53]Mikkola, P., Martin, O. A., Chandramouli, S., Hartmann, M., Pla, O. A., Thomas, O., Pesonen, H., Corander, J., Vehtari, A., Kaski, S., Bür̈kner, P.-C. and Klami, A. (2021) Prior knowledge elicitation: The past, present, and future.arxiv.org/abs/2112.01380
  51. [54]Murphy, K. P. (2022) Probabilistic Machine Learning: An introduction. MIT Press. URL: http://probml.github.io/book1.
  52. [55]Neal, R. M. (2011) MCMC using Hamiltonian dynamics. URL: http://arxiv.org/abs/1206.1901.
  53. [56]Orihara, S., Momozaki, T. and Sugasawa, S. (2026) Bayesian doubly robust causal inference via posterior coupling. URL: https://arxiv.org/abs/2506.04868.
  54. [57]Pakman, A. and Paninski, L. (2014) Exact Hamiltonian Monte Carlo for Truncated Multivariate Gaussians. Journal of Computational and Graphical Statistics, 23, 518–542. URL: http://www.jstor.org/stable/43305741.arxiv.org/abs/1208.4118
  55. [58]Parikh, N. and Boyd, S. (2014) Proximal algorithms. Found. Trends Optim., 1, 127–239. URL: https://doi.org/10.1561/2400000003.
  56. [59]Pas, S. L. v. d. (2021) Theoretical guarantees for the horseshoe and other global-local shrinkage priors. In Handbook of Bayesian Variable Selection, 133–160. Chapman and Hall/CRC.DOI
  57. [60]Pas, S. L. v. d., Kleijn, B. J. K. and Vaart, A. W. v. d. (2014) The horseshoe estimator: Posterior concentration around nearly black vectors. Electronic Journal of Statistics, 8, 2585 – 2618. URL: https://doi.org/10.1214/14-EJS962.
  58. [61]Pas, S. L. v. d., Salomond, J.-B. and Schmidt-Hieber, J. (2016) Conditions for posterior contraction in the sparse normal means problem. Electronic Journal of Statistics, 10, 976 – 1000. URL: https://doi.org/10.1214/16-EJS1130.
  59. [62]Pas, S. L. v. d., Szabó, B. and Vaart, A. W. v. d. (2017a) Adaptive posterior contraction rates for the horseshoe. Electronic Journal of Statistics, 11. URL: https://projecteuclid.org/journals/electronic-journal-of-statistics/volume-11/issue-2/Adaptive-posterior-contraction-rates-for-the-horseshoe/10.1214/17-EJS1316.full.arxiv.org/abs/1702.03698
  60. [64]Pereyra, M. (2016) Proximal Markov chain Monte Carlo algorithms. Statistics and Computing, 26, 745–760.
  61. [65]Petrov, V. V. (1975) Sums of Independent Random Variables, vol. 82 of Ergebnisse der Mathematik und ihrer Grenzgebiete. Berlin, Heidelberg: Springer-Verlag. URL: https://doi.org/10.1007/978-3-642-65809-9.
  62. [66]Piironen, J. and Vehtari, A. (2017) Sparsity information and regularization in the horseshoe and other shrinkage priors. Electronic Journal of Statistics, 11, 5018 – 5051. URL: https://doi.org/10.1214/17-EJS1337SI.
  63. [67]Polson, N. G. and Scott, J. G. (2011) Shrink Globally, Act Locally: Sparse Bayesian Regularization and Prediction*. In Bayesian Statistics 9 (eds. J. M. Bernardo, M. J. Bayarri, J. O. Berger, A. P. Dawid, D. Heckerman, A. F. M. Smith and M. West), 501–538. Oxford University Press. URL: https://academic.oup.com/book/1879/chapter/141655378.DOI
  64. [69]Polson, N. G., Scott, J. G. and Willard, B. T. (2015) Proximal algorithms in statistics and machine learning. URL: https://arxiv.org/abs/1502.03175.
  65. [70]Presman, R. and Xu, J. (2023) Distance-to-set priors and constrained bayesian inference. In International Conference on Artificial Intelligence and Statistics, 2310–2326. PMLR.arxiv.org/abs/2210.12258
  66. [71]Riha, A. E., Lindgren, L., Kohns, D., Bürkner, P.-C. and Vehtari, A. (2026) To select or not to select: predictively consistent priors instead of model selection. arXiv preprint arXiv:2606.22850.arxiv.org/abs/2606.22850
  67. [72]Robert, C. (2007) The Bayesian Choice: From Decision-Theoretic Foundations to Computational Implementation. Springer Texts in Statistics. Springer New York.
  68. [73]Schielzeth, H. (2010) Simple means to improve the interpretability of regression coefficients. Methods in Ecology and Evolution, 1, 103–113.DOI
  69. [74]Sewell, D. K. (2025) Posterior shrinkage towards linear subspaces. Bayesian Analysis, 20, 657–680.arxiv.org/abs/2401.07820
  70. [75]Shukla, A., Vats, D. and Chi, E. C. (2025) Proximal Hamiltonian Monte Carlo. arXiv preprint arXiv:2510.22252.arxiv.org/abs/2510.22252
  71. [76]Sivula, T., Magnusson, M., Matamoros, A. A. and Vehtari, A. (2025) Uncertainty in bayesian leave-one-out cross-validation based model comparison. Bayesian Analysis, 1, 1–31.
  72. [77]Stan Development Team (2026) Stan Modeling Language Users Guide and Reference Manual, Version 2.38. URL: http://mc-stan.org/.
  73. [78]Tadesse, M. and Vannucci, M. (2021) Handbook of Bayesian Variable Selection. Chapman & Hall/CRC Handbooks of Modern Statistical Methods. CRC Press.DOI
  74. [79]Vehtari, A. and Ojanen, J. (2012) A survey of Bayesian predictive methods for model assessment, selection and comparison. Statistics Surveys, 6, 142 – 228. URL: https://doi.org/10.1214/12-SS102.
  75. [80]Vehtari, A., Simpson, D., Gelman, A., Yao, Y. and Gabry, J. (2022) Pareto Smoothed Importance Sampling.
  76. [81]Wainwright, M. J. (2019) High-Dimensional Statistics: A Non-Asymptotic Viewpoint. Cambridge University Press.
  77. [82]West, M. (1987) On scale mixtures of normal distributions. Biometrika, 74, 646–648. URL: https://doi.org/10.1093/biomet/74.3.646.
  78. [84]Yanchenko, E. (2025) Pseudo-R2D2 prior for high-dimensional ordinal regression. Statistics and Computing, 35, 129.arxiv.org/abs/2502.17491
  79. [85]Yanchenko, E., Bondell, H. D. and Reich, B. J. (2024) Spatial regression modeling via the R2D2 framework. Environmetrics, 35, e2829.arxiv.org/abs/2301.09951
  80. [87]Yanchenko, E., Irie, K. and Sugasawa, S. (2026) The Group R2D2 Shrinkage Prior for Sparse Linear Models with Grouped Covariates. Statistics and Computing, 36, 57.arxiv.org/abs/2412.15293
  81. [88]Zellner, A. (1986) On assessing prior distributions and Bayesian regression analysis with g-prior distributions. Bayesian inference and decision techniques.
  82. [89]Zhang, Y. D., Naughton, B. P., Bondell, H. D. and Reich, B. J. (2022) Bayesian regression using a prior on the model fit: The R2-D2 shrinkage prior. Journal of the American Statistical Association, 117, 862–874.
  83. [90]Zhou, X., Heng, Q., Chi, E. C. and Zhou, H. (2024) Proximal MCMC for Bayesian inference of constrained and regularized estimation. The American Statistician, 78, 379–390.DOI

Paper details

vlnc:W_gt6kuzw4
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 9

downloads · 2

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents