New truthful calibration errors improve model ranking in multiclass prediction.
problem Non-truthful calibration errors can mislead model comparisons.
method Introduced perfectly truthful calibration errors for multiclass predictions.
result Truthful calibration errors preserve decision-theoretic dominance and stabilize model rankings.
Post-processing predictors reduces calibration errors for decision-making.
problem Predictors with low calibration error for machine learning may have high error for decision-making.
method Post-processing with ε distance to calibration adds noise to make predictions differentially private.
result Post-processing achieves O(√ε) ECE and CDL, asymptotically optimal.
Optimizes calibration error estimators for better classifier trustworthiness.
problem Lack of guidance on selecting and tuning calibration error estimators.
method Reformulates calibration estimation as a regression problem with i.i.d. input pairs.
result Demonstrates the effectiveness of optimized calibration estimators on image classification tasks.
New decision-theoretic calibration error metric improves prediction reliability.
problem Improving the reliability of predictions for decision-making.
method Proposed Calibration Decision Loss (CDL) and an efficient algorithm to achieve near-optimal CDL.
result Near-optimal CDL guarantees vanishing payoff loss from miscalibration.
Proposes a new calibration error estimator for deep neural networks.
problem Improves calibration of deep neural networks, especially for canonical calibration.
method Uses a Dirichlet kernel density estimate to create a low-bias, trainable calibration error estimator.
result Asymptotically converges to true Lp calibration error, enabling efficient estimation and mini-batch updates. Estimates proper calibration errors and refinement terms in probabilistic predictions.
problem Lack of a general estimator for proper calibration errors and refinement terms with known statistical properties.
method Proposes a method for consistent, asymptotically unbiased estimation of proper calibration errors and refinement terms.
result Proves the relation between refinement and f-divergences, implying information monotonicity in neural networks.
We extend a variational framework to estimate calibration errors for Lp divergences.
problem Ensuring predicted probabilities match observed class frequencies in machine learning.
method Extend variational framework to Lp divergences, separating over- and under-confidence. result Avoids overestimation and separates over- and under-confidence.
Optimal multiclass U-calibration error found to be Θ(√KT).
problem Online multiclass U-calibration with low regret for all bounded proper losses.
method Follow-the-Perturbed-Leader algorithm and lower bound construction.
result Optimal U-calibration error is Θ(√KT).
Simple algorithm achieves distance to calibration error of at most 2√T+1.
problem Achieving distance to calibration error of O(√T) in adversarial setting.
method An extremely simple, efficient, deterministic algorithm.
result Obtains distance to calibration error at most 2√T+1.
Simple proof shows forecasts can be calibrated in a few periods.
problem Ensuring forecasts are calibrated over multiple periods.
method Uses minimax theorem to prove existence and guarantees calibration error.
result Calibration can be achieved in N3 periods with error at most 1/N. Improves model calibration for deep neural networks using proper scores.
problem Calibration errors in deep neural networks are often biased and inconsistent.
method Introduces proper calibration errors related to proper scores.
result Demonstrates the superiority of proper scores over common estimators.
TCE measures calibration error with a test-based approach.
problem Measuring calibration error of probabilistic binary classifiers.
method TCE uses a novel loss function based on a statistical test.
result TCE offers clear interpretation, consistent scale, and enhanced visual representation.
A new calibration metric bridges testability and actionability.
problem Combining testability and actionable insights for forecast probabilities.
method Cutoff Calibration Error (CCE) that assesses calibration over intervals of forecasted probabilities.
result Cutoff Calibration Error is both testable and actionable.
New study on neural network calibration, linking it to generalization gap.
problem Neural networks lack strong guarantees on calibration.
method Decomposed calibration error into train set and generalization gap.
result Models with small generalization gap are well-calibrated.
Applications such as weather forecasting and personalized medicine demand models that output calibrated probability estimates---those representative of the true likelihood of a prediction. Most models are not calibrated out of the box but are recalibrated by post-processing model outputs. We find in this work that popu…
In safety-critical applications a probabilistic model is usually required to be calibrated, i.e., to capture the uncertainty of its predictions accurately. In multi-class classification, calibration of the most confident predictions only is often not sufficient. We propose and study calibration measures for multi-class…
We propose a new framework to improve the calibration of neural networks.
problem Improving the accuracy of model confidence predictions.
method Introducing a differentiable surrogate for expected calibration error (DECE) and a meta-learning framework to optimise model hyper-parameters for validation set calibration.
result Achieved competitive performance with existing calibration approaches.
Develops confidence intervals for ECE, a measure of model calibration.
problem Ensuring the calibration of probabilistic predictions in machine learning models.
method Develops confidence intervals for the ℓ2 Expected Calibration Error (ECE), considering top-1-to-k calibration. result Shows asymptotic normality and different convergence rates for calibrated and miscalibrated models, developing methods to construct valid confidence intervals.
Reassesses calibration metrics in machine learning models.
problem Inconsistent reporting of calibration metrics in recent literature.
method Calibration-based decomposition of Bregman divergences, visualization of calibration and generalization error.
result New visualization technique for detecting trade-offs between calibration and generalization.
A new metric CKCE improves model calibration comparison.
problem Comparing the calibration of probabilistic models is challenging.
method CKCE based on Hilbert-Schmidt norm of conditional mean operators.
result CKCE provides more consistent and robust model calibration comparisons.
Decision makers increasingly rely on algorithmic risk scores to determine access to binary treatments including bail, loans, and medical interventions. In these settings, we reconcile two fairness criteria that were previously shown to be in conflict: calibration and error rate equality. In particular, we derive necess…
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.
New calibration measure SCDL improves trust in AI predictions.
problem Improving trust in AI predictions by ensuring they are both actionable and testable.
method Introducing SCDL, a new calibration measure that is fully actionable and testable.
result SCDL is the first calibration measure that is fully actionable and testable.
This work proves L2-regularized ERM controls smCE without post-hoc correction.
problem Calibration of predicted probabilities in machine learning models.
method Canonical L2-regularized empirical risk minimization. result Theoretical proof that smCE is controlled by ERM without post-hoc correction.
CRUDE calibrates regression uncertainty without assuming specific error distributions.
problem Uncalibrated uncertainty estimates in regression models, especially for modern predictive tasks.
method CRUDE assumes error distributions have a constant shape, shifted and scaled by predicted mean and standard deviation.
result CRUDE produces sharper, better calibrated, and more accurate uncertainty estimates than existing methods.
The paper addresses poor calibration in fine-tuned LLMs after preference alignment.
problem Poor calibration in fine-tuned Large Language Models (LLMs) after preference alignment.
method Proposes a calibration-aware fine-tuning approach to restore calibration without compromising model performance.
result Demonstrates the effectiveness of the proposed methods through extensive experiments.
New method calibrates heterogeneous treatment effect models.
problem Difficulty in estimating and calibrating heterogeneous treatment effects.
method Defined and proposed a robust estimator for HTE calibration, based on doubly robust treatment effect estimators.
result Proposed method evaluates calibration of learned HTE models, addressing overfitting and high-dimensionality.
New framework for evaluating multiclass classifier calibration.
problem Ensuring classifiers are well-calibrated for trustworthy predictions.
method Utility Calibration framework that measures calibration error relative to a utility function.
result Unified and robust interpretation of existing calibration metrics.
PITMonitor monitors model calibration over time with formal error guarantees.
problem Fixed-sample tests applied to models over time can lead to false alarms.
method PITMonitor uses mixture e-processes to detect distributional shifts in probability integral transforms.
result PITMonitor achieves competitive detection rates on river's FriedmanDrift benchmark.
Certified calibration methods protect model confidence from adversarial attacks.
problem Adversarial attacks degrade model calibration, reducing confidence in predictions.
method Developed certified calibration methods to provide worst-case bounds on calibration under adversarial perturbations.
result Certified calibration methods produce analytic and approximate bounds for the Brier score and expected calibration error.
New methods reduce bias in estimating calibration error.
problem Reducing bias in estimating calibration error.
method Synthesizing model outputs and using equal-mass bins.
result Two reliable calibration-error estimators found: debiased estimator and ECE_sweep.
A new method calibrates value predictions in offline RL to improve reliability.
problem Difficulty in long-horizon value prediction in offline reinforcement learning.
method Bellman calibration, a weak reliability criterion, and Iterated Bellman Calibration.
result Finite-sample guarantees show that Bellman calibration error is controlled at nonparametric rates.
New method calibrates asynchronous, error-prone covariates for longitudinal data.
problem Estimation biases and slow convergence in analyzing time-varying covariates with measurement error.
method Functional calibration approach based on functional principal component analysis.
result Asymptotically unbiased and consistent estimators for time-invariant coefficients; optimal convergence rate for time-varying coefficients.
This post introduces model calibration and evaluation measures, highlighting issues with a common measure.
problem Ensuring model confidence accurately reflects true outcomes.
method Explains common calibration definition, ECE, and its drawbacks.
result New evaluation measures needed for comprehensive model calibration.
New method calibrates probabilistic regression models without restrictive assumptions.
problem Ensuring predictive distributions accurately reflect true uncertainty.
method Nonparametric re-calibration algorithm based on conditional kernel mean embeddings.
result Consistently outperforms prior re-calibration approaches across various benchmarks.
Improved pricing of vanilla options using modified Adams method and sinh-acceleration.
problem Calibration of rough Heston model leads to incorrect implied volatility surfaces.
method Modified Adams method and sinh-acceleration for Fourier inversion.
result Corrected implied volatility surface is significantly flatter and fits data poorly.
Paper proposes a method to predict MOBA game winners with calibrated confidence.
problem Predicting MOBA game winners with noisy data and uncertain noise.
method A novel confidence-calibration method considering data uncertainty.
result Achieves outstanding expected calibration error (ECE) of 0.57%.
The paper explores various forms of calibration scores and their implications for fairness.
problem The evaluation of probabilistic predictions through calibration.
method The authors organize three grouping choices and one agglomeration of group errors, providing a framework for comparing and creating new calibration scores.
result The study demonstrates that appropriate choices of grouping can provide notions of (sub-)group or individual fairness.
Paper proves a new lower bound on calibration error for binary prediction.
problem Proving a strong lower bound on calibration error for binary prediction.
method Developed two new techniques: early stopping and sidestepping.
result Proves an Ω(T0.528) lower bound on calibration error. Estimates calibration error under label shift without labels.
problem Ensuring model reliability in the face of dataset shift without access to labels.
method Importance re-weighting of the labeled source distribution to estimate calibration error under label shift.
result Effective and reliable CE estimation with respect to the shifted target distribution.
Study validates ML-UQ calibration statistics using simulated reference values.
problem Validation of ML-UQ calibration statistics is lacking due to lack of predefined reference values.
method Proposed validation workflow using simulated reference values derived from synthetic datasets.
result Some statistics, like CC and ENCE, are overly sensitive to generative distribution choice.
The paper provides a uniform convergence bound for smooth calibration error and its relationship with functional gradient.
problem Limited theoretical understanding of learning algorithms achieving high accuracy and good calibration.
method Focuses on smooth calibration error, providing a uniform convergence bound and proving the relationship with functional gradient.
result Derives conditions for simultaneous classification and calibration guarantees in gradient boosting trees, kernel boosting, and neural networks.
New calibration measure SSCE ensures truthful prediction, unlike existing measures.
problem Ensuring truthful calibration measures in sequential prediction.
method Introduced a new calibration measure, Subsampled Smooth Calibration Error (SSCE).
result SSCE ensures truthful prediction, while existing measures are far from truthful.
New method calibrates classifier probabilities with guaranteed coverage.
problem Inaccurate probability estimates by classifiers in high-risk applications.
method Adaptive temperature scaling algorithm for conformal prediction.
result Improves calibration error measures and standard metrics across various tasks.
We show how to compute the Bayes error-rate for speaker verifiers.
problem How many errors does a speaker verifier make in a hundred trials?
method We compute the Bayes error-rate using calibrated likelihood ratios and user-supplied prior probabilities.
result The Bayes error-rate is upper bounded by the minimum of EER, P, and 1-P.
Paper estimates optimal classification error with soft labels and calibration.
problem Estimating the optimal classification error with soft labels and calibration.
method Extends previous work on soft labels to estimate Bayes error, addressing bias and corrupted labels.
result The method provides a statistically consistent estimator of the Bayes error, even with imperfectly calibrated soft labels.
New algorithms adaptively calibrate predictions in non-stationary environments, matching optimal rates.
problem Designing online prediction algorithms that adapt to varying levels of non-stationarity.
method Epoch-based scheduling and non-uniform partitioning of the prediction space.
result Achieves adaptive calibration guarantees under multiple measures with optimal rates.
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.