PosCal training improves classification models by calibrating posterior probabilities.
problem Poorly calibrated posterior probabilities in classification models.
method End-to-end training procedure that directly optimizes the objective while minimizing the difference between predicted and empirical posterior probabilities.
result PosCal training achieves about 2.5% task performance gain and 16.1% calibration error reduction.
This paper argues against using calibration metrics for assessing posterior probabilities and proposes expected proper scoring rules instead.
problem The assessment of posterior probabilities generated by machine learning classifiers using calibration metrics is flawed and should be replaced with expected proper scoring rules.
method The paper reviews proper scoring rules from a practical perspective, explains why expected PSRs are a principled measure of posterior quality, and introduces a new calibration metric called calibration loss.
result Calibration loss is superior to expected calibration error and expected score divergence calibration metrics for assessing posterior probabilities.
Loss-calibrated EP improves Bayesian decision-making by focusing on utility-sensitive posterior approximations.
problem Bayesian decision-making under asymmetric utility functions.
method Loss-calibrated expectation propagation (Loss-EP) that tilts the posterior towards higher utility decisions.
result Loss-EP can capture useful information for decision-making under asymmetric penalties.
New method calibrates neural SBI to avoid overconfident posteriors.
problem Overconfident posteriors in SBI due to inaccurate uncertainty quantification.
method Introduces a calibration term into neural model training objective, enabling end-to-end backpropagation.
result Achieves competitive or better coverage and posterior density than existing methods.
Improved Bayesian FL method calibrates predictions for federated learning.
problem Overconfident predictions in Bayesian FL methods for federated learning.
method β-Predictive Bayes algorithm interpolates between mixture and product of local predictive posteriors, tuning parameter β for better calibration.
result Demonstrated superior calibration compared to other baselines, even with increased data heterogeneity.
Conformal Bayes under label shift: post-hoc calibration vs. in-training adaptation
problem Bayesian prediction sets under label shift
method Post-hoc calibration vs. In-training adaptation
result Both strategies achieve valid coverage equally in an unbiased training regime
CP4SBI improves the calibration of credible sets in SBI models.
problem Inaccurate credible sets in SBI models lead to underestimation of true parameters.
method Develops a local conformal calibration framework for SBI models.
result Improves the quality of uncertainty quantification for neural posterior estimators.
Calibrated probabilistic solvers improve accuracy of ODE estimates.
problem Uncertainty in probabilistic ODE solutions is not well-calibrated for adaptive step sizes.
method Introduce and assess several calibration methods for probabilistic ODE solvers.
result Calibration methods interact efficiently with adaptive step-size selection, improving posteriors.
FMCPE improves SBI accuracy by correcting posterior estimators with flow matching.
problem Model misspecification in SBI leads to biased or overconfident posteriors.
method Flow Matching Corrected Posterior Estimation (FMCPE) trains a posterior approximator and corrects it using calibration samples.
result FMCPE consistently mitigates misspecification effects, improving inference accuracy and uncertainty quantification.
New method improves calibration of BayesCG for better uncertainty quantification.
problem Bayesian conjugate gradient method's poor calibration limits its utility.
method Randomized postiteration strategy to enhance posterior calibration.
result The method improves the distribution of posterior errors and enhances uncertainty quantification.
Implied posterior probability of a given model (say, Support Vector Machines (SVM)) at a point x is an estimate of the class posterior probability pertaining to the class of functions of the model applied to a given dataset. It can be regarded as a score (or estimate) for the true posterior probability, which ca…
The paper corrects Bayesian neural network approximations to improve decision quality.
problem Inaccurate posterior approximations in Bayesian neural networks lead to suboptimal decisions.
method Develops methods to calibrate approximate posterior predictive distributions for better decision making.
result Empirically produces higher quality decisions compared to previous methods.
In this paper we study the probabilistic properties of the posteriors in a speech recognition system that uses a deep neural network (DNN) for acoustic modeling. We do this by reducing Kaldi's DNN shared pdf-id posteriors to phone likelihoods, and using test set forced alignments to evaluate these using a calibration s…
Two methods improve Gaussian process predictive distributions' calibration.
problem Improving the reliability of Gaussian process predictive intervals.
method Introduces two methods: cps-gp and bcr-gp, both adapting conformal predictive systems to GP interpolation.
result Both methods provide finite-sample marginal calibration and smooth predictive distributions.
A new method calibrates Gaussian processes for more accurate uncertainty estimates.
problem Uncertainty estimates from Gaussian processes are often miscalibrated in practice.
method A novel calibration approach using different hyperparameters to generate more accurate predictive quantiles.
result The method yields tighter predictive quantiles and is more flexible than existing approaches.
Blade uses diffusion priors to accurately and calibratedly infer complex systems.
problem Derivative-free Bayesian inversion for high-dimensional, nonlinear problems with costly forward models.
method Blade employs an ensemble of interacting particles and diffusion models as priors, querying forward models only through evaluations.
result Blade produces well-calibrated posterior samples that existing methods cannot, improving with more iterations and particles.
This paper establishes the asymptotic consistency of the {\it loss-calibrated variational Bayes} (LCVB) method. LCVB was proposed in~\cite{LaSiGh2011} as a method for approximately computing Bayesian posteriors in a `loss aware' manner. This methodology is also highly relevant in general data-driven decision-making con…
Two approaches improve conformal Bayes for label shift, one post-hoc and one in-training.
problem Improving prediction sets for target domain under label shift.
method Two complementary approaches: post-hoc calibration and in-training adaptation.
result In-training adaptation achieves up to 43% width reduction at unchanged coverage.
New PAC-Bayes meta-learning method improves few-shot learning accuracy and calibration.
problem Few-shot learning with limited data.
method PAC-Bayes framework extended to meta-learning, estimating task-specific posteriors.
result State-of-the-art calibration and classification results on benchmarks.
The semantic map calibrates uncertainty from language model probabilities.
problem Uncertainty in language model probabilities for professional decisions.
method Prespecified semantic map linking probabilities of verbal responses to probabilities of declared states.
result Language-derived probabilities outperform printed numerical probabilities and recover valid uncertainty coverage.
New research shows calibration error is flawed when dealing with model uncertainty.
problem Current model evaluation techniques conflate model uncertainty with aleatoric uncertainty.
method Posterior predictive checks to evaluate deep learning models.
result Calibration error and variants are incorrect when model uncertainty is present.
A new method improves SNPE for intractable likelihood models.
problem Simulation-based models with intractable likelihoods.
method Adaptive calibration kernel and variance reduction techniques.
result The proposed method provides a better approximation of the posterior.
Proposes a method to calibrate data for more accurate linear correlation testing.
problem Inaccurate Pearson's correlation coefficient due to sample size and data non-normality.
method Predictive data calibration using machine learning to condition data on expected linear relationship.
result Calibrated Pearson's correlation coefficient yields a calibrated p-value and r estimate for posterior probability interpretation.
New method estimates grouping loss in neural networks to improve confidence scores.
problem Improving confidence scores in neural networks to reflect true posterior probabilities.
method Proposed an estimator to approximate the grouping loss.
result Modern neural networks exhibit grouping loss, especially in distribution shifts.
Leaky ReLU activations improve the calibration of Bayesian neural networks.
problem Bayesian neural networks struggle with mean-field variational inference for ReLU activations.
method Investigated the effect of activation functions on the calibration of Bayesian neural networks.
result Leaky ReLU activations lead to more Gaussian-like weight posteriors and lower expected calibration error.
FMP sampling improves model calibration without sharing data.
problem Difficulties in specifying priors for modern neural networks.
method FMP sampling replaces prior and likelihood with a predictive distribution, running the sampler centrally.
result FMP sampling significantly improves calibration over baseline methods.
Posterior SBC validates inference conditionally on observed data.
problem Validating inference for specific observed data.
method Simulation-based calibration checking (SBC) adapted to use posterior parameters.
result Validates inference conditionally on observed data.
Bayesian decision theory outlines a rigorous framework for making optimal decisions based on maximizing expected utility over a model posterior. However, practitioners often do not have access to the full posterior and resort to approximate inference strategies. In such cases, taking the eventual decision-making task i…
Bayesian method improves approximate model posteriors.
problem Poor uncertainty quantification in approximate Bayesian inference.
method Optimizing a transformation of the approximate posterior to maximize a scoring rule.
result Significant reduction in bias and improvement in posterior coverage properties.
Bayesian models for networks are often misspecified, leading to overconfident inference.
problem Real-world networks violate assumptions of geometry and link function in latent space models.
method Proposes a generalized posterior framework for random geometric graphs, using Link-Sequential R-SafeBayes to adaptively tune posterior regularization.
result Improved calibration and better link prediction performance demonstrated on synthetic and real-world networks.
Focal loss improves classification but not class-posterior probability estimation.
problem Improving class-posterior probability estimation from focal loss.
method Proved classification-calibration and derived a transformation to recover true class-posterior probabilities.
result A transformation of the confidence score from focal loss minimization allows recovery of true class-posterior probabilities.
New algorithm samples Bayesian neural networks for improved calibration.
problem Improving calibration of Bayesian neural networks.
method Symmetric Minibatch Splitting-UBU (SMS-UBU) algorithm.
result SMS-UBU provides better calibration performance than standard methods.
New method samples from LLM posterior for coherent, useful responses.
problem Hallucinations in large language models.
method Posterior sampling for conditional generation, with calibration.
result Achieves statistical guarantees with higher downstream utility.
Misspecification-Aware Simulation-Based Inference via Side-Channel Guidance
problem Simulation-based inference (SBI) of latent parameters is hindered by simulator misspecification.
method Misspecification-Aware Simulation-Based Inference (MA-SBI) turns side-channel text into a posterior correction.
result MA-SBI matches the oracle posterior across 10 seeds and two backbones.
PBI inference may not be calibrated if predictive model is inaccurate.
problem Uncertainty quantification in PBI may be unreliable if the predictive model is not accurate.
method Predictive Bayesian inference with a forward predictive model.
result Posterior concentration depends on the predictive model, leading to potential calibration issues.
Proposes FIPO-BC for efficient online calibration of complex models.
problem Efficiently calibrating computationally expensive models with large datasets.
method Fixed inducing points online Bayesian calibration (FIPO-BC) algorithm.
result FIPO-BC is at least ten times faster than standard methods and enables online updates.
Bayesian neural SDEs calibrate financial models robustly.
problem Calibrating financial models using neural SDEs for robustness.
method Bayesian framework with prior and likelihood, global approximation theorem, Langevin algorithm.
result Robust bounds on implied volatility surface learned from historical and option data.
Bayesian calibration of black-box computer models offers an established framework to obtain a posterior distribution over model parameters. Traditional Bayesian calibration involves the emulation of the computer model and an additive model discrepancy term using Gaussian processes; inference is then carried out using M…
New GP-based method improves uncertainty quantification for causal functions.
problem Challenges in quantifying uncertainty for causal effects, especially for entire functions.
method GP-based approach using inner-product of observational functions in RKHS, with tractable posterior moments and calibration.
result Improves uncertainty quantification while maintaining causal effect estimation performance.
The study quantifies decision-making risks from suboptimal classifiers and proposes methods to reduce these risks.
problem Excess risk in decision-making from suboptimal probabilistic classifiers.
method Analytical expressions and upper/lower bounds for excess risk, calibration curve estimation, grouping loss estimator.
result Identifies regimes where recalibration alone or post-training is more effective.
Study shows TAP free energy minimization provides better posterior inference in high-dimensional linear models.
problem Deviation from true posterior mean and underestimation of posterior uncertainty in variational inference.
method Minimization of TAP free energy in a high-dimensional asymptotic framework, showing geometric and statistical properties.
result Local minimizer of TAP free energy provides consistent estimate of posterior marginals and correctly calibrated posterior inference.
New methods for tuning alpha in Gibbs posteriors improve speed and accuracy.
problem Inconsistency in Bayesian inference and lack of fast tuning methods for alpha.
method Proposed two data-driven methods: sample-splitting and bootstrapping. Formulated alpha-posteriors for three models.
result Sample-splitting outperforms SafeBayes in speed and accuracy, especially in complex models.
BayCANN uses ANN to speed up Bayesian calibration in health sciences.
problem Bayesian calibration's practical and computational burdens in health decision sciences.
method BayCANN trains an ANN metamodel to calibrate parameters probabilistically, comparing accuracy and speed to direct Bayesian calibration.
result BayCANN is more accurate and faster than direct Bayesian calibration methods.
LPF provides formal guarantees for aggregating multi-evidence in probabilistic tasks.
problem Lack of formal guarantees for multi-evidence reasoning in AI.
method LPF uses variational autoencoders and Sum-Product Networks to aggregate evidence items.
result Proves multiple formal guarantees including calibration preservation and error decay.
Meta-Cal improves post-hoc calibration of neural networks.
problem Improving the accuracy of uncalibrated neural network predictions.
method Meta-Cal uses a base calibrator and a ranking model with constraints to provide high-probability bounds.
result Meta-Cal significantly outperforms existing methods in post-hoc multi-class classification calibration.
In this paper we revisit the weighted likelihood bootstrap, a method that generates samples from an approximate Bayesian posterior of a parametric model. We show that the same method can be derived, without approximation, under a Bayesian nonparametric model with the parameter of interest defined as minimising an expec…
Bayesian neural networks improve uncertainty calibration without sacrificing accuracy.
problem Bayesian neural networks struggle with uncertainty calibration and high-dimensional geometry.
method Model uncertainty only in weight directions using a von Mises-Fisher posterior on the unit sphere, deriving a compact KL term.
result A lightweight, dimension-aware variational unit improves calibration without sacrificing accuracy.
New method corrects selection bias in complex models.
problem Selection bias in statistical studies leading to systematic distortions.
method Amortized Bayesian inference with neural posterior estimation.
result Recover well-calibrated posterior distributions across diverse selection mechanisms.