Calibrating classifiers reduces grouping loss using sufficiency criteria.
problem Grouping loss in probabilistic classifier calibration is often overlooked.
method Revisited Langford & Zadrozny's probing reduction approach and introduced Brier curves.
result The probing reduction approach reduces grouping loss and supports sufficient calibration.
Study uses Lie group subgroups to identify special subspaces in calibrations.
problem Identifying calibrated subspaces in Lie groups.
method Utilizes the principal three-dimensional subgroup of a simple Lie group.
result Identifies certain special subspaces as calibrated for invariant forms.
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.
Study shows pooling scores for conformal prediction distorts group coverage.
problem Pooling scores for conformal prediction distorts group coverage.
method Derived conservation law and lower bound, demonstrated tension between fairness definitions, quantified trade-off between policies.
result Pooling scores for conformal prediction distorts group coverage.
Improves model calibration and selection in unsupervised domain adaptation.
problem Distribution shifts in unsupervised domain adaptation.
method Developed a novel importance weighted group accuracy estimator.
result Improves state-of-the-art performances by 22% in model calibration and 14% in model selection.
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.
The paper addresses fairness issues in screening classifiers, proposing within-group monotonicity to avoid unfair treatment of qualified candidates.
problem Within-group unfairness in screening classifiers using calibrated models.
method Introducing within-group monotonicity as a property to avoid unfair treatment and developing an efficient post-processing algorithm based on dynamic programming.
result Within-group monotonicity can be achieved efficiently and often at a small cost, improving fairness without significantly compromising prediction accuracy.
We clarify what fairness guarantees we can and cannot expect to follow from unconstrained machine learning. Specifically, we characterize when unconstrained learning on its own implies group calibration, that is, the outcome variable is conditionally independent of group membership given the score. We show that under r…
GC-FCP provides efficient federated CP with group-conditional coverage guarantees.
problem Uncertainty quantification in federated settings with distributed calibration data.
method Group-conditional federated conformal prediction (GC-FCP) using group-stratified coresets.
result GC-FCP offers efficient aggregation and calibration compared to centralized methods.
This work addresses building fair and calibrated models.
problem Building models that are both fair and calibrated.
method Developed a new definition of fairness and showed that group-wise calibration results in fairness. Proposed post-processing techniques and modifications of calibration losses.
result Demonstrated that ensuring group-wise calibration results in a fair model under the new definition of fairness.
SOCP uses SOM to find groups and local calibration buffers for better regional coverage.
problem Heterogeneous regional coverage gaps in conformal prediction.
method Self-Organizing Map (SOM) for group discovery; local calibration buffers at BMU or fixed grid.
result Reduces regional coverage gaps on 7/8 benchmarks by 7.1%.
The study extends calibrated geometry to smooth maps and finds energy bounds.
problem Finding energy bounds for smooth maps between Riemannian manifolds.
method Generalizing calibrated submanifolds to smooth maps and applying to energy functional.
result Lower bounds to the energy of smooth maps in homotopy classes.
Adaptive uncertainty quantification improves black-box model predictions in generative AI.
problem Improving uncertainty quantification for black-box models in generative AI.
method Adaptive partitioning and local calibration of conformity scores.
result Local tightening of uncertainty sets with adaptive bands.
Paper proves tight lower bounds for online multicalibration, separating it from marginal calibration.
problem Proving lower bounds for online multicalibration in relation to marginal calibration.
method Information-theoretic approach, constructing group families from orthonormal bases.
result Establishes tight lower bounds for online multicalibration, matching upper bounds up to logarithmic factors.
LoBoost improves local conformal prediction for gradient-boosted trees without extra data splits.
problem Quantifying uncertainty in gradient-boosted tree predictions.
method Model-native local conformal prediction using leaf structure.
result Competitive interval quality and improved test MSE with large calibration speedups.
We investigate the deformation theory of a class of generalized calibrations in Riemannian manifolds for which the tangent bundle has reduced structure group U(n), SU(n), G_2 and Spin(7). For this we use the property of the associated calibration form to be parallel with respect to a metric connection which may have no…
Solves Plateau's Problem in Heisenberg group for graphs.
problem Plateau's Problem in the Heisenberg group for intrinsic graphs.
method Geometric construction and calibration argument.
result Solves Plateau's Problem under smallness conditions.
Unified framework for generalized Venn and Venn-Abers calibration for reliable prediction.
problem Asymptotic guarantees of popular distribution-free methods in model calibration.
method Unified framework extending Vovk's approach to generic loss functions, transforming predictors into set-valued predictions.
result Finite-sample set predictions shrink to a single conditionally calibrated prediction, capturing epistemic uncertainty.
Study aggregation of statistical evidence under unknown dependence using group-invariance.
problem Aggregating statistical evidence under unknown and complex dependence structures.
method Develops a framework using group-invariance and permutation-based constructions to aggregate evidence across transformed datasets.
result Shows uniform improvement in critical values for single-batch aggregation over deterministic calibrations, adapting to unknown dependence structures.
The machine learning community has become increasingly concerned with the potential for bias and discrimination in predictive models. This has motivated a growing line of work on what it means for a classification procedure to be "fair." In this paper, we investigate the tension between minimizing error disparity acros…
The exceptional holonomy groups are G2 in 7 dimensions, and Spin(7) in 8 dimensions. Riemannian manifolds with these holonomy groups are Ricci-flat. This is a survey paper on exceptional holonomy, in two parts. Part I introduces the exceptional holonomy groups, and explains constructions for compact 7- and 8-manifolds …
CCI combines Bayesian and gradient boosting to create fair, reliable credit risk scores.
problem Tackles high-stakes lending decisions with changing data distributions and fairness constraints.
method Combines Bayesian neural risk scorer and fairness-constrained gradient boosting with shift-aware fusion.
result CCI achieves best trade-off between discrimination, calibration, stability, and fairness.
Develops a method to estimate uncertainty for group-level recommendations in matrix completion.
problem Uncertainty estimation for group-level recommendations in matrix completion.
method Structured conformal inference method combining any matrix completion algorithm.
result Stronger group-level guarantees through structured calibration.
A popular methodology for building binary decision-making classifiers in the presence of imperfect information is to first construct a non-binary "scoring" classifier that is calibrated over all protected groups, and then to post-process this score to obtain a binary decision. We study the feasibility of achieving vari…
The paper reconciles two conflicting fairness criteria in algorithmic risk scores.
problem How to reconcile calibration and equal error rates in algorithmic risk scores.
method Derive necessary and sufficient conditions for existence of calibrated scores achieving equal error rates, then present an algorithm to find the most accurate score subject to both criteria.
result The method can eliminate error disparities while maintaining calibration and improve profit in credit lending.
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.
A novel approach models rating transitions using Lie groups and Deep Learning.
problem Modeling rating transitions with geometric properties and stochastic processes.
method Introducing Itô-SDEs on Lie groups, using TimeGAN for calibration, and examining rating matrix properties.
result The geometric approach using Lie groups and Deep Learning generates a good fit for rating transitions.
Study shows label errors impact model disparity metrics, proposing mitigation methods.
problem Impact of label errors on model disparity metrics.
method Empirical study, characterizing label error effects; proposing estimation and relabeling methods.
result Label errors significantly affect model disparity metrics, particularly for minority groups.
We mathematically compare four competing definitions of group-level nondiscrimination: demographic parity, equalized odds, predictive parity, and calibration. Using the theoretical framework of Friedler et al., we study the properties of each definition under various worldviews, which are assumptions about how, if at a…
This paper reviews metrics to assess AI model calibration accuracy.
problem AI model probabilities do not always match their true accuracy.
method Comprehensive review of 82 probability calibration metrics.
result Identified 4 classifier families and 1 object detection family of metrics.
LCMQR improves prediction intervals by adapting to local heteroscedasticity.
problem Efficient and adaptive prediction intervals for local heteroscedasticity.
method LCMQR combines multi-quantile information with kernel-based localization.
result LCMQR constructs tighter intervals than prior methods, especially in heterogeneous environments.
We study massless deformations of generalized calibrated cycles, which describe, in the language of generalized complex geometry, supersymmetric D-branes in N=1 supersymmetric compactifications with fluxes. We find that the deformations are classified by the first cohomology group of a Lie algebroid canonically associa…
Algorithm improves binary classification of biased grouped data.
problem Improving binary classification for biased, grouped data.
method Assumes partition-projected class-conditional invariance across groups and derives a semi-supervised algorithm to learn a group-aware classifier.
result Demonstrates improved area under the ROC curve compared to baselines.
We study locally conformal calibrated G2-structures whose underlying Riemannian metric is Einstein, showing that in the compact case the scalar curvature cannot be positive. As a consequence, a compact homogeneous 7-manifold cannot admit an invariant Einstein locally conformal calibrated G2-structure unless the…
Paper develops a new algorithm to improve screening processes.
problem Improving the efficiency and accuracy of screening processes.
method Developed a distribution-free screening algorithm called Calibrated Subset Selection (CSS).
result CSS provides near-optimal shortlists with provable diversity guarantees.
The paper constructs calibrated submanifolds in Euclidean spaces with specific symmetries.
problem Finding calibrated submanifolds in Euclidean spaces with given symmetries.
method Constructing submanifolds invariant under Lie group actions and using specific ansatzes.
result Explicitly determined special Lagrangian submanifolds and rigidity results.
Recommender systems are personalized: we expect the results given to a particular user to reflect that user's preferences. Some researchers have studied the notion of calibration, how well recommendations match users' stated preferences, and bias disparity the extent to which mis-calibration affects different user grou…
A new framework evaluates LLM calibration in open-ended QA.
problem Evaluating LLM calibration in open-ended QA settings.
method Sem-ECE framework: sampling answers, grouping by semantic classes, and using frequencies as confidence.
result Sem-ECE estimators are unbiased and Sem2 achieves smaller calibration error. Many classification applications require accurate probability estimates in addition to good class separation but often classifiers are designed focusing only on the latter. Calibration is the process of improving probability estimates by post-processing but commonly used calibration algorithms work poorly on small data…
New deep Cox mixture model improves survival analysis performance.
problem Challenges in survival analysis due to censoring and healthcare applications.
method Learning mixtures of Cox regressions with deep neural networks for hazard ratios and non-parametric baseline hazard.
result Our approach outperforms classical and modern survival analysis methods, especially in minority demographics.
The paper studies harmonic graphs in the Heisenberg group and their properties.
problem No analogous theorem exists for H-minimal surfaces in the Heisenberg group. method Introduced intrinsic Dirichlet energy and studied its critical points (contact harmonic graphs).
result Calibration condition and construction of energy-minimizing graphs with various singularities.
We construct a compact formal 7-manifold with a closed G2-structure and with first Betti number b1=1, which does not admit any torsion-free G2-structure, that is, it does not admit any G2-structure such that the holonomy group of the associated metric is a subgroup of G2. We also construct associative ca…
The t copula is often used in risk management as it allows for modelling tail dependence between risks and it is simple to simulate and calibrate. However, the use of a standard t copula is often criticized due to its restriction of having a single parameter for the degrees of freedom (dof) that may limit its capabilit…
New insights link no-regret learning to online conformal prediction in adversarial settings.
problem Understanding the relationship between no-regret learning and online conformal prediction in adversarial environments.
method Analysis of existing algorithms and new connections between no-regret learning and conformal prediction.
result No-regret learning algorithms can provide group-conditional coverage guarantees in adversarial settings.
Estimates KL divergence with fairness considerations for sub-populations.
problem Fairly estimate KL divergence between distributions considering sub-populations.
method Proposes multi-group attribution for KL divergence estimation, derived from multi-calibration.
result Shows multi-group attribution provides better KL divergence estimates conditioned on sub-populations.
The study examines when to trust confidence thresholding in pseudo-labelling regression.
problem Calibrated probabilities from classifiers used for pseudo-labelling need careful handling to avoid bias in downstream regression.
method Developed a diagnostic apparatus to predict and bound the bias induced by confidence thresholding, derived a closed-form expression for the attenuation bias.
result The bias can be predicted from the residual score variance V∗, motivating a structural separation between classifier features and downstream controls. Benchmark assesses fairness in algorithmic uncertainty, revealing consistent and calibrated estimates improve fairness.
problem Challenges in managing uncertainty in fairness evaluations for predictive algorithms.
method Introduces FairlyUncertain, an axiomatic benchmark for evaluating uncertainty in fairness.
result Consistent and calibrated uncertainty estimates improve fairness without explicit fairness interventions.
MACI improves LLM factuality inference with higher retention and lower time cost.
problem Ensuring factuality in LLM responses for high-stakes domains.
method Reformulated conformal inference in a multiplicative filtering setting, leveraging ensembles for more accurate factuality scores and group-conditional calibration.
result MACI achieves higher retention and lower time cost compared to baselines, preserving validity through group-conditional calibration.