T-Cal tests model calibration with a minimax optimal test.
problem Detecting mis-calibration of predictive models using a finite validation dataset.
method T-Cal is a minimax optimal test for calibration based on a debiased plug-in estimator of the ℓ2-Expected Calibration Error (ECE). result T-Cal is a practical tool for testing the calibration of probabilistic classification methods.
New test assesses probabilistic model calibration without expensive approximations.
problem Assessing calibration of probabilistic models with scores.
method Kernel Calibration Conditional Stein Discrepancy (KCCSD) test using new score-based kernels.
result Control over type-I error with improved scalability and efficiency.
New algorithm tests model calibration in nearly-linear time.
problem Testing model calibration from samples efficiently.
method Reformulated as minimum-cost flow, solved with dynamic programming.
result Optimal testing problem solved in nearly-linear time.
New calibration bands for various distributions improve testing for auto-calibration.
problem Testing for auto-calibration in finite samples is challenging.
method Construct calibration bands for the exponential dispersion family using finite sample properties.
result Calibration bands allow for various tests for calibration and auto-calibration.
The paper introduces a spline-based method for calibrating neural networks.
problem Ensuring neural network outputs are reliable for safety-critical applications.
method Approximating the empirical cumulative distribution function using splines to map network outputs to calibrated probabilities.
result The spline-based recalibration consistently outperforms existing methods on calibration measures.
ECCIT improves conditional independence tests by calibrating for miscalibration.
problem Inaccurate frequentist guarantees in CITs, especially in small samples and misspecified models.
method Empirically Calibrated Conditional Independence Tests (ECCIT) that optimize and correct for miscalibration.
result ECCIT achieves valid FDR with higher power than existing calibration strategies.
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.
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.
Calibration without labels in multiple testing
problem Interpretable error probabilities in large-scale hypothesis testing
method Constructing pseudo-labels from spacings of ordered p-values result Finding that q-value can be severely miscalibrated Statistical test verifies long-term rating system calibration with overlapping time windows.
problem Verifying supervisory requirements for overlapping time windows in rating systems.
method Analyzes long-run default rate distribution and correlation effects; presents conservative calibration test methods.
result Developed a test for individual and portfolio levels that can handle unknown variance.
New method detects strong calibration in ML models, even for small poorly calibrated subgroups.
problem Auditing machine learning models for strong calibration is difficult, especially for small poorly calibrated subgroups.
method Reorder observations by expected residuals and use changepoint detection for score-based cumulative sum (CUSUM) test.
result The proposed adaptive CUSUM test consistently achieved higher power and more than doubled power in auditing mortality risk prediction models.
Paper proposes a unified framework for evaluating calibration of probabilistic models.
problem Evaluation of calibration for general probabilistic predictive models.
method Unified framework for calibration evaluation and tests for any probabilistic model.
result Generalization and reformulation of existing measures and tests.
Boosting trees can test necessary conditions for regression model calibration.
problem Testing calibration and auto-calibration in regression models.
method Using boosting trees to test calibration and auto-calibration.
result Boosting trees prove to be very powerful in testing calibration and auto-calibration in large insurance datasets.
The paper extends calibration to sets of probabilistic classifiers, finding many ensembles are poorly calibrated.
problem Evaluating the validity of epistemic uncertainty in sets of probabilistic classifiers.
method Proposed a novel nonparametric calibration test for sets of probabilistic classifiers.
result Ensembles of deep neural networks are often not well calibrated.
New method improves model calibration by adjusting confidence based on prediction correctness.
problem Improving model confidence alignment with true class probabilities.
method Post-hoc calibration objective using transformed samples for training.
result Competitive calibration performance on in-distribution and out-of-distribution test sets.
A framework assesses the trustworthiness of probabilistic classifiers using local calibration error.
problem Assessing the trustworthiness of probabilistic classifiers beyond traditional metrics.
method I-trustworthy framework linking local calibration to trustworthiness; Kernel Local Calibration Error (KLCE) method for hypothesis testing.
result The effectiveness of the proposed test statistic demonstrated through simulated and real-world datasets.
Ensemble models improve prediction calibration for mismatched distributions.
problem Calibration issues in deep neural networks with mismatched train and test distributions.
method Simple data augmentation and mixing techniques for ensemble models.
result Improves calibration and accuracy on CIFAR10 and CIFAR100 benchmarks.
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…
Paper proposes an efficient method for calibrating spatio-temporal forecasts.
problem Real-world spatio-temporal forecasting challenges like signal anomalies and distributional shifts.
method Learning with Calibration (ST-TTC) for real-time bias correction.
result ST-TTC improves spatio-temporal forecasting accuracy with reduced computational cost.
The paper tackles confidence calibration for exploratory machine learning problems.
problem Difficulty in curating datasets and confusion about category validity.
method Introduces four new algorithms for category-specific confidence estimation, including kernel density ratios.
result Kernel density ratios provide a novel approach to confidence calibration, especially for exploratory problems.
A test assesses the calibration of set-based epistemic uncertainty representations.
problem Evaluating the accuracy of set-based representations of epistemic uncertainty in machine learning.
method Proposes a novel statistical test to determine if a convex combination of predictions is calibrated, allowing instance-level variability.
result Demonstrates the benefits of capturing instance-level variability on synthetic and real-world experiments.
The uncertainty estimation is critical in real-world decision making applications, especially when distributional shift between the training and test data are prevalent. Many calibration methods in the literature have been proposed to improve the predictive uncertainty of DNNs which are generally not well-calibrated. H…
Framework calibrates ML models for risk control in various tasks.
problem Achieving statistical guarantees for model predictions.
method Reframing risk control as multiple hypothesis testing, applying statistical techniques.
result New calibration methods for multi-label classification, instance segmentation, outlier detection, and confidence set coverage.
This paper tackles over-certainty in test-time adaptation models, proposing a solution to improve calibration.
problem Over-certainty in predictions caused by domain shifts, leading to misplaced trust.
method Introduces a certainty regularizer that dynamically adjusts pseudo-label confidence based on backbone entropy and logit norm.
result Achieves state-of-the-art performance in terms of Expected Calibration Error and Negative Log Likelihood, while maintaining accuracy.
ORCA calibrates LLMs for efficient, generalizable reasoning.
problem Miscalibration of large language models leading to inefficiencies.
method Online Reasoning Calibration (ORCA) using conformal prediction and test-time training.
result ORCA provides higher efficiency and generalization across different reasoning tasks.
SGD-trained models' disagreement predicts test error.
problem Estimating test error of deep networks.
method Empirical testing and theoretical analysis of SGD ensembles.
result SGD ensembles' disagreement correlates with test error.
Study robustness of split conformal prediction under adversarial attacks.
problem Ensuring distribution-free coverage guarantees in CP under adversarial conditions.
method Theoretical analysis and extensive experiments on split conformal prediction robustness.
result Prediction coverage varies with calibration-time attack strength, enabling control over coverage under adversarial tests.
Paper develops a new method to improve model calibration under distribution shifts.
problem Challenges in uncertainty quantification with different training and test distributions.
method Develops multi-domain temperature scaling to handle distribution shifts.
result Outperforms existing methods on in-distribution and out-of-distribution test sets.
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…
PD curve calibration refers to the transformation of a set of rating grade level probabilities of default (PDs) to another average PD level that is determined by a change of the underlying portfolio-wide PD. This paper presents a framework that allows to explore a variety of calibration approaches and the conditions un…
New method uses interval-based metric to validate prediction uncertainty in machine learning.
problem Validation of prediction uncertainty in machine learning regression tasks is unreliable due to heavy-tailed distributions.
method Shift from variance-based metrics to interval-based Prediction Interval Coverage Probability (PICP).
result PICP method more quickly and reliably tests prediction intervals than variance-based metrics.
Proposes a new method to improve Bayesian computation accuracy using flexible classification.
problem Bayesian computations accuracy check using rank-based simulation-based calibration has limitations.
method Replaces marginal rank test with a flexible classification approach that learns from data.
result Improves statistical power and provides an interpretable divergence measure of miscalibration.
A post-hoc framework improves model performance by calibrating different feature spaces.
problem Improving AUC performance on binary classification tasks for overconfident models.
method Identifies heterogeneous partitions of the feature space and applies post-hoc calibration techniques to each partition.
result Theoretical optimality of the framework for any model, demonstrated on deep neural networks.
Corrects bias in LLM-as-a-judge evaluations using adaptive calibration.
problem Bias in LLM evaluations due to imperfect sensitivity and specificity.
method Plug-in framework with confidence intervals accounting for test and calibration dataset uncertainties.
result LML-based evaluation yields more reliable estimates than human-only evaluation.
Semi-supervised method boosts two-sample testing with covariate data.
problem Two-sample testing with covariate information.
method Semi-supervised kernel test with asymptotic normality.
result Higher asymptotic power compared to existing methods.
Fast nonparametric conditional independence testing via two-stage regression
problem Fast nonparametric conditional independence testing
method BLITZ (Broad-to-Local Independence Testing via residualiZation)
result Better null calibration than fast kernel, random-feature, and regression-based competitors
Framework for fast CAT calibration and administration using AutoML and IRT.
problem Calibrating and administering large-scale CAT tests with limited data.
method AutoIRT (AutoML + IRT) for calibration, BanditCAT for administration.
result Framework successfully launched new item types on DET practice test.
Proposes MCLLO for assessing and recalibrating multiclass probability predictions.
problem Limited multicategory recalibration methods for assessing and comparing model calibration.
method MCLLO recalibration method that assesses calibration without model access and is easy to interpret.
result MCLLO outperforms other methods in simulations and real-world case studies.
Efficiently calibrates Bergomi models to VIX derivatives using vector quantization.
problem Calibrating Bergomi models to VIX derivatives for accurate pricing.
method Applied vector quantization in mixed Bergomi models for fast and efficient option pricing.
result Calibration of Bergomi models to VIX derivatives is feasible and accurate over daily data.
Study compares various calibration methods for binary classification tasks.
problem Improving probabilistic predictions in binary classification models.
method Benchmarked 21 classifiers using 5 calibration methods on real data.
result Venn-Abers predictors and Beta calibration show the largest log-loss reductions.
BYOV combines SSL and Bayesian methods for uncertainty estimation.
problem Model uncertainty in applications.
method Combines Bootstrap Your Own Latent (BYOL) and Bayes by Backprop (BBB).
result BYOV improves model calibration and reliability with various augmentations.
Deep neural networks (DNNs) are poorly calibrated when trained in conventional ways. To improve confidence calibration of DNNs, we propose a novel training method, distance-based learning from errors (DBLE). DBLE bases its confidence estimation on distances in the representation space. In DBLE, we first adapt prototypi…
Improves survival prediction model calibration for better individual decision-making.
problem Survival prediction's marginal and conditional calibration issues.
method Conformal prediction using individual survival probabilities.
result Effective marginal and conditional calibration without compromising discrimination.
We present a scoring approach for speaker verification that mimics the standard PLDA-based backend process used in most current speaker verification systems. However, unlike the standard backends, all parameters of the model are jointly trained to optimize the binary cross-entropy for the speaker verification task. We …
CP improves robustness against distribution shift using physics-informed structural causal models.
problem Uncertainty in machine learning predictions under distributional shift.
method Physics-informed structural causal model (PI-SCM) to upper bound coverage difference.
result PI-SCM improves coverage robustness across confidence levels and test domains.
A new method calibrates forecasts without sacrificing expertise.
problem Forecasters' calibration scores can be manipulated to appear expert.
method Deterministic and stochastic online procedures to calibrate forecasts.
result Calibration can be achieved without losing expertise.
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.
Improved active output selection reduces calibration time by 10% or more.
problem Efficiently calibrate models with noisy data.
method Improved active output selection strategy considering noise estimate.
result At least 10% fewer measurements needed compared to existing strategies.