Bayesian robustness to outliers in linear regression and ratio estimation

  title={Bayesian robustness to outliers in linear regression and ratio estimation},
  author={Alain Desgagn'e and Philippe Gagnon},
  journal={Brazilian Journal of Probability and Statistics},
Whole robustness is a nice property to have for statistical models. It implies that the impact of outliers gradually decreases to nothing as they converge towards plus or minus infinity. So far, the Bayesian literature provides results that ensure whole robustness for the location-scale model. In this paper, we make two contributions. First, we generalise the results to attain whole robustness in simple linear regression through the origin, which is a necessary step towards results for general… 

Figures and Tables from this paper

A New Bayesian Approach to Robustness Against Outliers in Linear Regression
This paper proposes a model with super heavy-tailed errors, and proves that it is wholly robust, meaning that the impact of outliers gradually vanishes as they move further and further away form the general trend.
An automatic robust Bayesian approach to principal component regression
A Bayesian approach that is robust to outliers in both the dependent variable and the covariates is introduced, compared to its nonrobust Bayesian counterpart, the traditional frequentist approach and a commonly employed robust frequentist method.
Robustness against conflicting prior information in regression
Including prior information about model parameters is a fundamental step of any Bayesian statistical analysis. It is viewed positively by some as it allows, among others, to quantitatively
Robust estimation of skew-normal distribution with location and scale parameters via log-regularly varying functions
The method of robust parameter estimation for location and scale parameters in the skew-normal distribution is proposed. One of the methods of rejection of outliers is the use of heavy-tailed
A Heavy-Tailed and Overdispersed Collective Risk Model
Insurance data can be asymmetric with heavy tails, causing inadequate adjustments of the usually applied models. To deal with this issue, hierarchical models for collective risk with heavy-tails of
Sélection de modèles robuste : régression linéaire et algorithme à sauts réversibles
The key result is the weak convergence of the sequence of stochastic processes engendered by the algorithm, which is, to the authors' knowledge, the first weak convergence result for the reversible jump algorithm.
A step further towards automatic and efficient reversible jump algorithms
This paper builds on the approach of Zanella (2019) for discrete spaces to incorporate information about neighbouring models and relies on approximations to posterior model probabilities that are asymptotically exact, as the sample size increases.
Informed reversible jump algorithms
This paper builds on the approach of Zanella (2020) for discrete spaces to incorporate information about neighbouring models and relies on approximations to posterior model probabilities that are asymptotically exact, proving that, in some scenarios, the samplers behave like those that use the exact model probabilities and sample from the parameter distributions, in the large sample regime.
Multi-Label Learning via Feature and Label Space Dimension Reduction
This paper proposes a new framework for multi-label learning with a large number of class labels and features, i.e., Multi-Label Learning via Feature and Label Space Dimension Reduction, namely MLL-FLSDR, in which both the feature space and label space are reduced to low dimensional spaces respectively.
Data analytics of social media publicity to enhance household waste management


Robustness to outliers in location–scale parameter model using log-regularly varying distributions
Estimating the location and scale parameters is common in statistics, using, for instance, the well-known sample mean and standard deviation. However, inference can be contaminated by the presence of
Bayesian Robustness Modelling of Location and Scale Parameters
Abstract.  The modelling process in Bayesian Statistics constitutes the fundamental stage of the analysis, since depending on the chosen probability laws the inferences may vary considerably. This is
Outlier Models and Prior Distributions in Bayesian Linear Regression
SUMMARY Bayesian inference in regression models is considered using heavy-tailed error distri- butions to accommodate outliers. The particular class of distributions that can be con- structed as
Outlier Resistant Alternatives to the Ratio Estimator
Abstract Many techniques have been suggested to lower the impact of outliers on sample survey estimates. Outliers can be downweighted by winsorization; that is, by replacing extreme data points by a
A class of methods for robust regression is developed, based on estimators of scale, that are introduced because of their invulnerability to large fractions of contaminated data and are proposed to be called “S-estimators”.
Outlier Robust Finite Population Estimation
Abstract Outliers in sample data are a perennial problem for applied survey statisticians. Moreover, it is a problem for which traditional sample survey theory offers no real solution, beyond the
A bayesian approach to some outlier problems.
The problem of outlying observations is considered from a Bayesian viewpoint and the linear model is considered, which assumes that a good observation is normally distributed about its mean with variance o.2, and a bad one is normal with the same mean but a larger variance.
A Fast Algorithm for S-Regression Estimates
This article proposes an analogous algorithm for computing S-estimators, that is based on a “local improvement” step of the resampling initial candidates, and performs a simulation study which shows that S-ESTimators computed with the fast-S algorithm compare favorably to the LTS-estimates computed withThe fast-LTS algorithm.
Bayesian heavy-tailed models and conflict resolution: A review
It is shown that Bayesian modelling with heavy-tailed distributions has been shown to produce more reasonable con‡ict resolution, typically by favouring one source of information over the other.