Visualization in Bayesian workflow

@article{Gabry2019VisualizationIB,
  title={Visualization in Bayesian workflow},
  author={Jonah Gabry and Daniel P. Simpson and Aki Vehtari and Michael Betancourt and Andrew Gelman},
  journal={Journal of the Royal Statistical Society: Series A (Statistics in Society)},
  year={2019}
}
Bayesian data analysis is about more than just computing a posterior distribution, and Bayesian visualization is about more than trace plots of Markov chains. Practical Bayesian data analysis, like all data analysis, is an iterative process of model building, inference, model checking and evaluation, and model expansion. Visualization is helpful in each of these stages of the Bayesian workflow and it is indispensable when drawing inferences from the types of modern, high-dimensional models that… Expand
ArviZ a unified library for exploratory analysis of Bayesian models in Python
While conceptually simple, Bayesian methods can be mathematically and numerically challenging. Probabilistic programming languages (PPLs) implement functions to easily build Bayesian models togetherExpand
Increasing Interpretability of Bayesian Probabilistic Programming Models Through Interactive Representations
TLDR
This work proposes the automatic transformation of Bayesian probabilistic models, expressed in a probabilism programming language, into an interactive graphical representation of the model's structure at varying levels of granularity, with seamless integration of uncertainty visualization. Expand
Bayesian statistics and modelling
TLDR
This Primer on Bayesian statistics summarizes the most important aspects of determining prior distributions, likelihood functions and posterior distributions, in addition to discussing different applications of the method across disciplines. Expand
Lumen: A software for the interactive visualization of probabilistic models together with data
TLDR
As the main feature of Lumen a user can rapidly and incrementally build flexible and potentially complex interactive visualizations of both the probabilistic model and the data that the model was trained on. Expand
Improving Bayesian Statistics Understanding in the Age of Big Data With the Bayesvl R Package
TLDR
The bayesvl R package is an open program, designed for implementing Bayesian modeling and analysis using the Stan language’s no-U-turn (NUTS) sampler, that can improve the user experience and intuitive understanding when constructing and analyzing Bayesian network models. Expand
Designing for Interactive Exploratory Data Analysis Requires Theories of Graphical Inference
TLDR
It is described how without a grounding in theories of human statistical inference, research in exploratory visual analysis can lead to contradictory interface objectives and representations of uncertainty that can discourage users from drawing valid inferences. Expand
Bayesian Data Analysis in Empirical Software Engineering Research
TLDR
This paper presents Bayesian data analysis techniques that work better on the same data---as they can provide clearer results that are simultaneously robust and nuanced, and demonstrates concrete advantages of using Bayesian techniques. Expand
Choosing priors in Bayesian ecological models by simulating from the prior predictive distribution
TLDR
A workflow for prior selection is demonstrated using simulation and visualization with two ecological examples and it is suggested that this difficulty can be overcome by simulating from the prior predictive distribution and visualizing the results on the scale of the response variable. Expand
Causal Support: Modeling Causal Inferences with Visualizations
TLDR
It is found that chart users’ causal inferences tend to be insensitive to sample size such that they deviate from the authors' normative benchmark, and on average users do not perform reliably better with common visualizations than they do with textual contingency tables. Expand
Graphical outputs and Spatial Cross-validation for the R-INLA package using INLAutils.
TLDR
This paper describes and illustrates the use of graphical summary plotting functions and the SLOOCV approach to compare the predictive performance of multiple spatial models and identifies the limits of the approach. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 35 REFERENCES
Exploratory Data Analysis for Complex Models
TLDR
This article proposes an approach to unify exploratory data analysis with more formal statistical methods based on probability models, developed in the context of examples from fields including psychology, medicine, and social science. Expand
Statistical inference for exploratory data analysis and model diagnostics
  • A. Buja, D. Cook, +4 authors H. Wickham
  • Medicine, Biology
  • Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences
  • 2009
TLDR
The proposed protocols will be useful for exploratory data analysis, with reference datasets simulated by using a null assumption that structure is absent, and teachers might find that incorporating these protocols into the curriculum improves their students’ statistical thinking. Expand
ggplot2 - Elegant Graphics for Data Analysis
This book describes ggplot2, a new data visualization package for R that uses the insights from Leland Wilkisons Grammar of Graphics to create a powerful and flexible system for creating dataExpand
Hamiltonian Monte Carlo for Hierarchical Models
Hierarchical modeling provides a framework for modeling the complex interactions typical of problems in applied statistics. By capturing these relationships, however, hierarchical models alsoExpand
The Prior Can Often Only Be Understood in the Context of the Likelihood
TLDR
This paper resolves an apparent paradox in prior modeling: a model encoding true prior information should be chosen without reference to the model of the measurement process, but almost all common prior modeling techniques are implicitly motivated by a reference likelihood. Expand
Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC
TLDR
An efficient computation of LOO is introduced using Pareto-smoothed importance sampling (PSIS), a new procedure for regularizing importance weights, and it is demonstrated that PSIS-LOO is more robust in the finite case with weak priors or influential observations. Expand
Pareto Smoothed Importance Sampling
Importance weighting is a general way to adjust Monte Carlo integration to account for draws from the wrong distribution, but the resulting estimate can be noisy when the importance ratios have aExpand
Yes, but Did It Work?: Evaluating Variational Inference
TLDR
Two diagnostic algorithms are proposed that give a goodness of fit measurement for joint distributions, while simultaneously improving the error in the estimate. Expand
Model Determination Using Predictive Distributions with Implementation via Sampling-Based Methods
TLDR
Model determination is divided into the issues of model adequacy and model selection and it is proposed to validate conditional predictive distributions arising from single point deletion against observed responses. Expand
Penalising Model Component Complexity: A Principled, Practical Approach to Constructing Priors
In this paper, we introduce a new concept for constructing prior distributions. We exploit the natural nested structure inherent to many model components, which defines the model component to be aExpand
...
1
2
3
4
...