Model patching closes subgroup performance gaps in skin cancer classification.
problem Inconsistent model performance on specific subgroups of a class.
method Two-stage framework that models subgroup features and learns semantic transformations, followed by data augmentation.
result Reductions in robust error of up to 33% relative to best baseline on benchmark datasets.
SHIFT framework identifies subgroups with large ML model performance decay.
problem Large model performance decay in subgroups when deployed.
method Subgroup-scanning Hierarchical Inference Framework (SHIFT) for performance drift.
result SHIFT identifies interpretable subgroups with large performance decay and suggests targeted actions to mitigate it.
A new method improves AI fairness assessment by estimating performance across intersectional subgroups.
problem Limited evaluation of AI systems across intersectional subgroups due to small sample sizes.
method Structured regression approach to disaggregated evaluation.
result Our method yields more accurate performance estimates, especially for small subgroups.
New framework shows diverse training data improves subgroup and overall performance.
problem Lack of understanding how diverse training data affects subgroup and overall performance.
method Casts data collection as part of the learning process, analyzes dataset compositions, and guides dataset design.
result Diverse representation in training data improves subgroup and overall performance.
The potential for learned models to amplify existing societal biases has been broadly recognized. Fairness-aware classifier constraints, which apply equality metrics of performance across subgroups defined on sensitive attributes such as race and gender, seek to rectify inequity but can yield non-uniform degradation in…
Causal Interaction Trees identify treatment subgroup effects in observational data.
problem Identifying subgroups with enhanced treatment effects in observational studies.
method Extending Classification and Regression Trees with subgroup-specific treatment effect estimators.
result The proposed algorithms enhance treatment effect heterogeneity in subgroups.
This paper evaluates fairness in deep metric learning and proposes a method to reduce subgroup performance gaps.
problem The negative impact of deep metric learning representations on minority subgroup performance in downstream tasks.
method Definition of fairness in DML through inter-class, intra-class, and uniformity properties; finDML benchmark; Partial Attribute De-correlation (PARADE) method.
result Bias in DML representations propagates to downstream tasks, even with balanced training data.
A new method clusters heterogeneous subgroups for accurate causal learning.
problem Diverse causal relationships across different time spans, regions, or strategies.
method Nonlinear Causal Kernel Clustering
result Reduction in prediction error through enhanced causal learning.
Selective regression allows abstention to improve fairness criteria.
problem Selective regression can exacerbate disparities between subgroups.
method Proposes new fairness criteria and two approaches to mitigate performance disparity.
result Proposed fairness criteria ensures performance improvement for every subgroup with reduced coverage.
In this paper we derive a generating series for the number of cellular complexes known as pavings or three-dimensional maps, on n darts, thus solving an analogue of Tutte's problem in dimension three. The generating series we derive also counts free subgroups of index n in $Δ^+ = \mathbb{Z}_2*\mathbb{Z}_2*\mathbb{Z…
The growing capability and accessibility of machine learning has led to its application to many real-world domains and data about people. Despite the benefits algorithmic systems may bring, models can reflect, inject, or exacerbate implicit and explicit societal biases into their outputs, disadvantaging certain demogra…
ADHAM provides interpretable survival analysis for healthcare.
problem Limited interpretability in deep learning survival models.
method Additive Deep Hazard Analysis Mixtures (ADHAM) with latent subgroup structure.
result ADHAM offers interpretable insights into exposure-outcome associations.
Paper proposes a federated learning framework for relative fairness.
problem Traditional fairness in federated learning overlooks performance disparities between client subgroups.
method Uses a minimax problem approach to minimize relative unfairness, introducing a fairness index based on loss ratios.
result Empirical evaluations confirm the framework's effectiveness in maintaining model performance while reducing disparity.
A statistical toolbox for analyzing model performance in medical imaging.
problem Analyzing model performance by patient and recording properties, especially in medical imaging.
method Selection of appropriate performance metrics, correction of multiple comparisons, and finding interesting subgroups.
result Enables rigorous assessment of model performance for potential subgroup disparities.
The paper tackles fairness in forecasting and learning linear dynamical systems.
problem Under-representation bias in training data for multiple subgroups.
method Introducing subgroup-fair and instant-fair learning of LDS from multiple trajectories of varying lengths, using hierarchies of convexifications of non-commutative polynomial optimisation problems.
result Empirical results show both the beneficial impact of fairness considerations on statistical performance and encouraging effects of exploiting sparsity on run time.
GAME improves matrix completion by considering subgroup-specific latent structures.
problem Heterogeneous data with overlapping categories, smoothing away subgroup-specific variation.
method Group-Aware Matrix Estimation (GAME) with overlapping nuclear-norm penalties.
result GAME outperforms global low-rank estimators in structured missingness regimes.
Posterior conformal prediction improves prediction interval validity for subgroups.
problem Marginal and conditional prediction interval validity for subgroups.
method Modeling conditional nonconformity score distribution as a mixture of cluster distributions.
result PCP produces tighter prediction intervals, especially for well-represented clusters.
AdaptHetero uses MLI to tailor EHR models for subgroup-specific predictions.
problem Lack of subgroup-specific, operationalizable modeling strategies in EHRs.
method Integrates MLI with unsupervised clustering to identify subgroup-specific characteristics.
result Improves predictive performance by up to 174.39 percent across many subpopulations.
New method reduces variance in subpopulation model performance estimates.
problem High variance in subpopulation performance metrics for small groups.
method Using an evaluation model to form model-based metric (MBM) estimates.
result MBMs produce more accurate and lower variance estimates for small subpopulations.
GLMM trees identify subgroups with different growth patterns in longitudinal data.
problem Identifying subgroups with distinct growth trajectories in longitudinal studies.
method Extended GLMM trees for longitudinal data.
result Extended GLMM trees outperform other methods in accuracy and speed.
CRE discovers interpretable subgroups with heterogeneous treatment effects.
problem Identifying subgroups with notable treatment effect heterogeneity.
method Causal Rule Ensemble (CRE) using an ensemble-of-trees approach.
result CRE offers interpretable decision rules and high stability in subgroup discovery.
In this paper, we investigate the effect of machine learning based anonymization on anomalous subgroup preservation. In particular, we train a binary classifier to discover the most anomalous subgroup in a dataset by maximizing the bias between the group's predicted odds ratio from the model and observed odds ratio fro…
D3M debiases models by selectively removing problematic examples.
problem Model failures on underrepresented subgroups.
method Isolates and removes specific training examples that cause failures.
result Efficiently trains debiased classifiers with minimal example removal.
In this paper, we compute the subgroup distortion of all finitely generated subgroups of all finitely generated 3-manifold groups, and the subgroup distortion in this case can only be linear, quadratic, exponential and double exponential. It turns out that the subgroup distortion of a subgroup of a 3-manifold group is …
Improves fairness in machine learning by adding underrepresented group data.
problem Machine learning biases across subgroups due to under-representation or societal biases.
method Data augmentation via pairwise mixup across subgroups to balance subpopulations.
result Achieves fair outcomes with robust if not improved accuracy.
The paper introduces a model to measure ASR fairness, addressing key issues.
problem Measuring fairness in ASR systems for different subgroups.
method Mixed-effects Poisson regression to control nuisance factors and handle unobserved heterogeneity.
result The method effectively addresses WER gaps among subgroups and is flexible for practical analyses.
Regular subgroups of SL3(R) are identified and ruled out.
problem Identifying and characterizing regular subgroups of SL3(R).
method Using Kapovich–Leeb–Porti and Guichard–Wienhard divergent subgroups criteria, and Oh's results.
result Regular subgroups of SL3(R) are precisely lattices in minimal horospherical subgroups.
Study on braid group quotients by congruence subgroups.
problem Understanding the image of congruence subgroups in GL(n,Z).
method Characterization through symplectic congruence subgroups.
result Open problem solved: image of congruence subgroups in GL(n,Z).
New framework for interpreting disaggregated fairness evaluations using causal models.
problem Misinterpretation of disaggregated fairness evaluations due to data representativeness and selection bias.
method Causal graphical models to characterize fairness properties and metric stability under different data generating processes.
result Disaggregated evaluations are unreliable without explicit assumptions regarding bias mechanisms.
The paper explores geometric finiteness in mapping class groups and constructs new examples of these subgroups.
problem Understanding geometric finiteness in mapping class groups and constructing new examples.
method Examined several constructions of subgroups and determined conditions for geometric finiteness.
result Provides new examples of parabolically geometrically finite and reducibly geometrically finite subgroups.
Proposes a new method for finding non-redundant, standout subgroups in numeric datasets.
problem Mining large numbers of redundant subgroups in numeric datasets.
method Dispersion-aware problem formulation based on MDL principle for subgroup set discovery.
result Empirically demonstrates SSD++ returns outstanding subgroup lists.
Proves Congruence Subgroup Property for two types of groups.
problem Proving Congruence Subgroup Property for specific groups.
method Elementary proof of Johnson filtration and geometric subsurface inclusions.
result Proves Congruence Subgroup Property for nilpotent quotients and subsurface subgroups.
New method constructs non-quasiconvex subgroups in hyperbolic groups.
problem Creating non-quasiconvex subgroups in hyperbolic groups.
method Using Stallings-like techniques on right-angled Coxeter groups (RACGs).
result Explicit examples of non-quasiconvex subgroups constructed.
Method provides statistical guarantees for identifying subgroups in ML studies.
problem Bias and noise in estimating conditional average treatment effects (CATE).
method Develops uniform confidence bands (GATES) for estimating group average treatment effects (GATEs).
result Identifies subgroups with statistical guarantees, regardless of effect size.
Characterizes knotted subgroups of Lie groups and provides examples.
problem Defining and understanding knotted subgroups of Lie groups.
method Geometric equivalence, one-parameter subgroups, infinitesimal elements, canonical forms, spectrum analysis.
result Completely classified knotted subgroups of SL(2,R) and SL(3,R).
Study subgroups of pro-p PD^3 groups, finding specific conditions.
problem Characterize subgroups of pro-p PD^3 groups. method Analyzes properties of subnormal and finitely presented subgroups.
result Conditions on subgroups of pro-p PD^3 groups. No hyperbolic group can have an infinite chain of free subgroups of fixed rank.
problem Infinite ascending chains of free subgroups in hyperbolic groups.
method Proof by contradiction and properties of hyperbolic groups.
result Hyperbolic groups do not contain strictly ascending chains of free quasiconvex subgroups of constant rank.
Robust subgroup discovery finds non-redundant, statistically significant subgroups.
problem Finding interpretable, robust subgroups from data.
method Formulated subgroup lists for univariate and multivariate targets, used MDL principle and greedy heuristic SSD++.
result SSD++ outperforms previous methods in quality and size of subgroup lists.
Sparse GFA identifies disease factors in FTD subgroups.
problem Heterogeneity in neurological disorders hinders understanding and treatment.
method Sparse Group Factor Analysis (GFA) with regularised horseshoe priors.
result Identified latent disease factors differentially expressed in FTD subgroups.
Let N be at least 4. We prove that every injective homomorphism from the Torelli subgroup into Out(FN) differs from the inclusion by a conjugation in Out(FN). This applies more generally to the following subgroups: every finite-index subgroup of Out(FN) (recovering a theorem of Farb and Handel); every subgro…
New lattices in higher dimensions have dense surface subgroups.
problem Finding dense subgroups in higher-dimensional arithmetic lattices.
method Exhibited nonuniform arithmetic lattices in SO(n,1).
result Contain Zariski-dense surface subgroups.
For a finitely generated group, there are two recent generalizations of the notion of a quasiconvex subgroup of a word-hyperbolic group, namely a stable subgroup and a Morse or strongly quasiconvex subgroup. Durham and Taylor defined stability and proved stability is equivalent to convex cocompactness in mapping class …
A new algorithm COVA-FC improves subgroup-fair clustering efficiency.
problem Challenges in making cluster assignments independent of sensitive attributes in subgroups.
method Defining a subgroup-fairness gap, deriving a covariance-based surrogate, and introducing a continuous relaxation for efficient optimization.
result COVA-FC achieves competitive cost-fairness trade-offs and improves computational efficiency.
Proposes a method to detect anomalies in multi-subgroup normal data.
problem Anomaly detection with limited labeled anomalies and multi-subgroup normal data.
method Learn multi-normal prototypes with deep embedding clustering and contrastive learning. Estimate the likelihood of unlabeled samples being normal during training.
result Superior performance compared to state-of-the-art methods on various datasets.
We associate cube complexes called completions to each subgroup of a right-angled Coxeter group (RACG). A completion characterizes many properties of the subgroup such as whether it is quasiconvex, normal, finite-index or torsion-free. We use completions to show that reflection subgroups are quasiconvex, as are one-end…
We show that, in an Artin-Tits group of spherical type, the intersection of two parabolic subgroups is a parabolic subgroup. Moreover, we show that the set of parabolic subgroups forms a lattice with respect to inclusion. This extends to all Artin-Tits groups of spherical type a result that was previously known for bra…
Study answers arithmeticity question for normal subgroup of lattices.
problem Arithmeticity of discrete subgroups of semisimple Lie groups with dense commensurators.
method Examined normal subgroups of lattices in semisimple Lie groups.
result Positive answer to Greenberg-Shalom's question for lattices.
Characterizes groups arising as fixed subgroups of RAAG automorphisms.
problem Identifying groups that can be fixed by finite-order automorphisms of RAAGs.
method Geometric characterisation using divisible cube complexes.
result Surface groups and commutator subgroups of RAAGs are fixed subgroups.