A menu-driven software package of Bayesian nonparametric (and parametric) mixed models for regression analysis and density estimation

Karabatsos, George

doi:10.3758/s13428-016-0711-7

Cited by 17 publications

(18 citation statements)

References 122 publications

(136 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…No such program exists for Bayesian regression, although we envision this as being the eventual outlet for Bayesian regression. Several exceptional scripts and programs have been written and are free to use (e.g., Karabatsos, 2017; Kruschke, 2011). Nonetheless, a logical and important next step is to develop more widely accessible, user-friendly interfaces for researchers and, pending much additional investigation, interventionists.…”

Section: Discussionmentioning

confidence: 99%

Bayesian asymmetric regression as a means to estimate and evaluate oral reading fluency slopes.

Solomon

Forsberg

2017

School Psychology Quarterly

View full text Add to dashboard Cite

Bayesian techniques have become increasingly present in the social sciences, fueled by advances in computer speed and the development of user-friendly software. In this paper, we forward the use of Bayesian Asymmetric Regression (BAR) to monitor intervention responsiveness when using Curriculum-Based Measurement (CBM) to assess oral reading fluency (ORF). An overview of Bayesian methods and their application to the problem-solving model is first presented, which is further illustrated by a case example. We conclude the paper with a Monte Carlo simulation study demonstrating the validity of BAR, as compared to the current standard of practice for CBM decision-making, ordinary least squares (OLS) regression. Results suggest that BAR is most advantageous with studies using small-to-moderate sample sizes, and when distributional information (such as the probability of intervention success) is of interest. (PsycINFO Database Record

show abstract

Section: Discussionmentioning

confidence: 99%

Bayesian asymmetric regression as a means to estimate and evaluate oral reading fluency slopes.

Solomon

Forsberg

2017

School Psychology Quarterly

View full text Add to dashboard Cite

show abstract

“…In this section, we discuss in detail the features and functionalities offered in three R packages addressing BNP density estimation, namely: BNPdensity, BNPmix (Canale et al, 2019), and DPpackage (Jara Karabatsos, 2017), or meta-analysis (bspmma, Burr, 2012) are not discussed here. Likewise, non Bayesian approaches are deliberately set aside.…”

Section: Package Comparisonmentioning

confidence: 99%

“…All three can be accessed directly from R by respectively using R2OpenBUGS/R2WinBUGS (Sturtz et al, 2005), rjags (Plummer, 2019), runjags (Denwood, 2016), and rstan (Stan Development Team, 2018). Programs for specific fields of Bayesian statistics have appeared in recent years, for instance bspmma (Burr, 2012) for meta-analysis using Dirichlet Process Mixture (DPM) models, DPpackage (Jara, 2007;Jara et al, 2011), a bundle of functions for Bayesian nonparametric models, BNPmix (Canale et al, 2019), a set of functions for density estimation with Dirichlet process and Pitman-Yor mixing measures via marginal algorithms, PReMiuM (Liverani et al, 2015) for profile regression using the Dirichlet process, Biips (Todeschini et al, 2014) for Bayesian inference via particle filtering, Bayesian Regression (Karabatsos, 2017) for Bayesian nonparametric regression. Packages mcclust (Scrucca et al, 2016), mcclust.ext (Wade and Ghahramani, 2018) and GreedyEPL (Rastelli and Friel, 2018) provide point estimation and credible sets for Bayesian cluster analysis.…”

mentioning

confidence: 99%

BNPdensity: Bayesian nonparametric mixture modeling in R

Arbel¹,

King²,

Lijoi³

et al. 2021

Preprint

View full text Add to dashboard Cite

Robust statistical data modelling under potential model mis-specification often requires leaving the parametric world for the nonparametric. In the latter, parameters are infinite dimensional objects such as functions, probability distributions or infinite vectors. In the Bayesian nonparametric approach, prior distributions are designed for these parameters, which provide a handle to manage the complexity of nonparametric models in practice. However, most modern Bayesian nonparametric models seem often out of reach to practitioners, as inference algorithms need careful design to deal with the infinite number of parameters. The aim of this work is to facilitate the journey by providing computational tools for Bayesian nonparametric inference. The article describes a set of functions available in the R package BNPdensity in order to carry out density estimation with an infinite mixture model, including all types of censored data. The package provides access to a large class of such models based on normalized random measures, which represent a generalization of the popular Dirichlet process mixture. One striking advantage of this generalization is that it offers much more robust priors on the number of clusters than the Dirichlet. Another crucial advantage is the complete flexibility in specifying the prior for the scale and location parameters of the clusters, because conjugacy is not required. Inference is performed using a theoretically grounded approximate sampling methodology known as the Ferguson & Klass algorithm. The package also offers several goodness of fit diagnostics such as QQ-plots, including a cross-validation criterion, the conditional predictive ordinate. The proposed methodology is illustrated on a classical ecological risk assessment method called the Species Sensitivity Distribution (SSD) problem, showcasing the benefits of the Bayesian nonparametric framework.

show abstract

“…2011), a bundle of functions for Bayesian nonparametric models, BNPmix (Canale, Corradin & Nipoti 2019), a set of functions for density estimation with Dirichlet process and Pitman–Yor mixing measures via marginal algorithms, PReMiuM (Liverani et al . 2015) for profile regression using the Dirichlet process, Biips (Todeschini, Caron & Fuentes 2014) for Bayesian inference via particle filtering, Bayesian Regression (Karabatsos 2017) for Bayesian nonparametric regression. Packages mcclust (Scrucca et al .…”

Section: Introductionmentioning

confidence: 99%

“…All three can be accessed directly from R by respectively using R2OpenBUGS/R2WinBUGS (Sturtz, Ligges & Gelman 2005), rjags (Plummer 2019), runjags (Denwood 2016), and rstan (Stan Development Team 2018). Programmes for specific fields of Bayesian statistics have appeared in recent years, for instance bspmma (Burr 2012) for meta-analysis using Dirichlet process mixture (DPM) models, DPpackage (Jara 2007;Jara et al 2011), a bundle of functions for Bayesian nonparametric models, BNPmix (Canale, Corradin & Nipoti 2019), a set of functions for density estimation with Dirichlet process and Pitman-Yor mixing measures via marginal algorithms, PReMiuM (Liverani et al 2015) for profile regression using the Dirichlet process, Biips (Todeschini, Caron & Fuentes 2014) for Bayesian inference via particle filtering, Bayesian Regression (Karabatsos 2017) for Bayesian nonparametric regression. Packages mcclust (Scrucca et al 2016), mcclust.ext (Wade & Ghahramani 2018) and GreedyEPL (Rastelli & Friel 2018) provide point estimation and credible sets for Bayesian cluster analysis.…”

mentioning

confidence: 99%

BNPdensity: Bayesian nonparametric mixture modelling in R

Arbel

King

Lijoi

et al. 2021

Aus NZ J of Statistics

View full text Add to dashboard Cite

Robust statistical data modelling under potential model mis-specification often requires leaving the parametric world for the nonparametric. In the latter, parameters are infinite dimensional objects such as functions, probability distributions or infinite vectors. In the Bayesian nonparametric approach, prior distributions are designed for these parameters, which provide a handle to manage the complexity of nonparametric models in practice. However, most modern Bayesian nonparametric models seem often out of reach to practitioners, as inference algorithms need careful design to deal with the infinite number of parameters. The aim of this work is to facilitate the journey by providing computational tools for Bayesian nonparametric inference. The article describes a set of functions available in the R package BNPdensity in order to carry out density estimation with an infinite mixture model, including all types of censored data. The package provides access to a large class of such models based on normalised random measures, which represent a generalisation of the popular Dirichlet process mixture. One striking advantage of this generalisation is that it offers much more robust priors on the number of clusters than the Dirichlet. Another crucial advantage is the complete flexibility in specifying the prior for the scale and location parameters of the clusters, because conjugacy is not required. Inference is performed using a theoretically grounded approximate sampling methodology known as the Ferguson & Klass algorithm. The package also offers several goodness-of-fit diagnostics such as QQ plots, including a cross-validation criterion, the conditional predictive ordinate. The proposed methodology is illustrated on a classical ecological risk assessment method called the species sensitivity distribution problem, showcasing the benefits of the Bayesian nonparametric framework.

show abstract

A menu-driven software package of Bayesian nonparametric (and parametric) mixed models for regression analysis and density estimation

Cited by 17 publications

References 122 publications

Bayesian asymmetric regression as a means to estimate and evaluate oral reading fluency slopes.

Bayesian asymmetric regression as a means to estimate and evaluate oral reading fluency slopes.

BNPdensity: Bayesian nonparametric mixture modeling in R

BNPdensity: Bayesian nonparametric mixture modelling in R

Contact Info

Product

Resources

About