Robust Bayesian inference for big data: combining sensor-based records with traditional survey data

DOI10.1214/21-AOAS1531zbMATH Open1498.62355arXiv2101.07456OpenAlexW3123085124MaRDI QIDQ2154206FDOQ2154206

Authors: Ali Rafei, Carol A. C. Flannagan, Brady T. West, Michael R. Elliott 

Publication date: 14 July 2022

Published in: The Annals of Applied Statistics (Search for Journal in Brave)

Abstract: Big Data often presents as massive non-probability samples. Not only is the selection mechanism often unknown, but larger data volume amplifies the relative contribution of selection bias to total error. Existing bias adjustment approaches assume that the conditional mean structures have been correctly specified for the selection indicator or key substantive measures. In the presence of a reference probability sample, these methods rely on a pseudo-likelihood method to account for the sampling weights of the reference sample, which is parametric in nature. Under a Bayesian framework, handling the sampling weights is an even bigger hurdle. To further protect against model misspecification, we expand the idea of double robustness such that more flexible non-parametric methods, as well as Bayesian models, can be used for prediction. In particular, we employ Bayesian additive regression trees, which not only capture non-linear associations automatically but permit direct quantification of the uncertainty of point estimates through its posterior predictive draws. We apply our method to sensor-based naturalistic driving data from the second Strategic Highway Research Program using the 2017 National Household Travel Survey as a benchmark.

Full work available at URL: https://arxiv.org/abs/2101.07456

Recommendations

• A survey of Bayesian statistical approaches for big data
• Data Integration by Combining Big Data and Survey Sample Data for Finite Population Inference
• Comment: A brief survey of the current state of play for Bayesian computation in data science at big-data scale
• Bayesian empirical likelihood inference with complex survey data
• Nonparametric Bayesian aggregation for massive data
• Bayesian data fusion: Probabilistic sensitivity analysis for unmeasured confounding using informative priors based on secondary data
• Robust Bayesian analysis: sensitivity to the prior
• Practical Bayesian modeling and inference for massive spatial data sets on modest computing environments^†

zbMATH Keywords

big dataprediction modeldoubly robustBayesian additive regression treesquasi-randomizationaugmented inverse propensity weightingnonprobability sample

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Applications of statistics to economics (62P20) Statistical aspects of big data and data science (62R07) Sampling theory, sample surveys (62D05)

Cites Work

• BART: Bayesian additive regression trees
• A Distributional Approach for Causal Inference Using Propensity Scores
• Demystifying double robustness: a comparison of alternative strategies for estimating a population mean from incomplete data
• Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models
• Title not available (Why is that?)
• Doubly Robust Estimation in Missing Data and Causal Inference Models
• Beta Regression for Modelling Rates and Proportions
• The central role of the propensity score in observational studies for causal effects
• Estimation of Regression Coefficients When Some Regressors Are Not Always Observed
• Resampling Inference With Complex Survey Data
• Model feedback in Bayesian propensity score estimation
• Estimation with missing data: beyond double robustness
• Doubly robust inference with missing data in survey sampling
• Using Calibration Weighting to Adjust for Nonignorable Unit Nonresponse
• To Model or Not To Model? Competing Modes of Inference for Finite Population Sampling
• On the Bias of the Multiple-Imputation Variance Estimator in Survey Sampling
• Parametric and semi-parametric estimation of regression models fitted to survey data
• Assessing lack of common support in causal inference using Bayesian nonparametrics: Implications for evaluating the effect of breastfeeding on children's cognitive outcomes
• Bayesian approach for addressing differential covariate measurement error in propensity score methods
• A two-step Bayesian approach for propensity score analysis: simulations and case study
• A comparative study of doubly robust estimators of the mean with missing data
• Modelling and generating correlated binary variables
• Combining data from two independent surveys: a model-assisted approach
• Inference for nonprobability samples
• Imputation using response probability
• Bayesian propensity score analysis for clustered observational data
• On the Validity of Inferences from Non-random Sample
• Statistical paradises and paradoxes in big data. I: Law of large populations, big data paradox, and the 2016 US presidential election
• Predicting human-driving behavior to help driverless vehicles drive: random intercept Bayesian additive regression trees
• Inference for domains under imputation for missing survey data
• Doubly Robust Inference With Nonprobability Survey Samples
• Robust Model-Based Inference for Incomplete Data via Penalized Spline Propensity Prediction
• A Note on Handling Nonresponse in Sample Surveys

Cited In (1)

• Bayesian estimation methods for survey data with potential applications to health disparities research

This page was built for publication: Robust Bayesian inference for big data: combining sensor-based records with traditional survey data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2154206)