Model collapse occurs in structured interactive learning if and only if the directed interaction graph satisfies a specific topological condition, with finite-sample guarantees for linear regression and asymptotic results for M-estimators.
Title resolution pending
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 8verdicts
UNVERDICTED 8roles
background 1polarities
background 1representative citing papers
A decision-theoretic parametric ROC framework under scale mixtures of skew-normal distributions defines optimal cutoffs by minimizing weighted misclassification risk, establishes their existence and uniqueness under monotone likelihood ratio, and supplies asymptotic normality with a plug-in variance
Generalized Wasserstein barycenters on Riemannian manifolds are absolutely continuous when all input measures are absolutely continuous, for strictly convex cost profiles h with singularity at zero, via a geometric approximation approach.
Defines threshold breakdown point and m-sensitivity for M-estimators, derives their properties, extends to hypothesis testing, and supplies consistency, asymptotic normality, and multiplier bootstrap results.
A new bootstrap goodness-of-fit test for the logistic propensity score model under nonignorable missing data, based on marginal sum-of-squared residuals, with asymptotic size and power guarantees.
MARS weights ranks in CD diagrams by the scaled distance between best and worst model performance to produce magnitude-sensitive statistical comparisons.
Numerical benchmarks indicate generative regularizers deliver strong reconstructions in some imaging inverse problem settings but can be unstable or problematic under imperfect conditions compared to variational methods.
A Gaussian mixture MIL framework with partially subsampled instances improves metastasis prediction accuracy on breast cancer whole-slide images over prior methods.
citing papers explorer
-
When Does Model Collapse Occur in Structured Interactive Learning?
Model collapse occurs in structured interactive learning if and only if the directed interaction graph satisfies a specific topological condition, with finite-sample guarantees for linear regression and asymptotic results for M-estimators.
-
Parametric ROC Analysis and Optimal Cutoff Selection under Scale Mixtures of Skew-Normal Distributions: A Decision-Theoretic Framework with Asymptotic Inference
A decision-theoretic parametric ROC framework under scale mixtures of skew-normal distributions defines optimal cutoffs by minimizing weighted misclassification risk, establishes their existence and uniqueness under monotone likelihood ratio, and supplies asymptotic normality with a plug-in variance
-
Absolute continuity of generalized Wasserstein barycenters of finitely many measures
Generalized Wasserstein barycenters on Riemannian manifolds are absolutely continuous when all input measures are absolutely continuous, for strictly convex cost profiles h with singularity at zero, via a geometric approximation approach.
-
The Threshold Breakdown Point
Defines threshold breakdown point and m-sensitivity for M-estimators, derives their properties, extends to hypothesis testing, and supplies consistency, asymptotic normality, and multiplier bootstrap results.
-
A goodness-of-fit test for the logistic propensity score model under nonignorable missing data
A new bootstrap goodness-of-fit test for the logistic propensity score model under nonignorable missing data, based on marginal sum-of-squared residuals, with asymptotic size and power guarantees.
-
MARS: Magnitude-Aware Rank Statistics
MARS weights ranks in CD diagrams by the scaled distance between best and worst model performance to produce magnitude-sensitive statistical comparisons.
-
A Stability Benchmark of Generative Regularizers for Inverse Problems
Numerical benchmarks indicate generative regularizers deliver strong reconstructions in some imaging inverse problem settings but can be unstable or problematic under imperfect conditions compared to variational methods.
-
Detecting Breast Carcinoma Metastasis on Whole-Slide Images by Partially Subsampled Multiple Instance Learning
A Gaussian mixture MIL framework with partially subsampled instances improves metastasis prediction accuracy on breast cancer whole-slide images over prior methods.