Multi-source transfer learning incurs an intrinsic adaptation cost that can exceed one, with phase transitions separating regimes where bias-agnostic estimators match oracle performance from those where they cannot.
hub
Population-level balance in signed networks
13 Pith papers cite this work. Polarity classification is still indexing.
13
Pith papers citing it
hub tools
verdicts
UNVERDICTED 13representative citing papers
Establishes statistical and computational optimality thresholds for common subspace estimation and inference under varying SNR regimes, including an impossibility result for adaptive confidence intervals below strong inference SNR.
Introduces a nonparametric inference procedure based on a sparse signed graphon model that yields valid confidence intervals for balance parameters and reports strong empirical evidence for balance theory across real signed networks.
A density-ratio framework compresses BMA posteriors into hard or soft support regions with explicit TV, KL, and predictive distortion bounds under predictor redundancy.
Develops TWSF estimator for causal forecasting in panel data by combining synthetic controls with time-series models under low-rank latent factor assumptions, providing finite-sample bounds and asymptotic normality.
A two-sample test for subspace equality in networks uses the Frobenius norm of projection matrix differences, with proven asymptotic normality to Gaussian under logarithmic average degree growth.
Mixed Poisson regression models with Gaussian latent variables are asymptotically robust to infinite target values but not to infinite covariate values, as shown for Poisson-Gamma, Poisson-log-t, and Poisson-RSB targets.
AI4BayesCode generates validated modular stateful MCMC samplers from natural language Bayesian model descriptions via LLM translation, modular blocks, and recursive stateful composition.
A distributionally robust ODE learning framework for heterogeneous systems that uses worst-case optimization over convex derivative combinations to produce a stabilized weighted estimator with theoretical guarantees.
StatsClaw is a multi-agent AI workflow that separates code implementation, data simulation, and testing via information barriers to produce faithful statistical software while keeping human control over methodological choices.
A double machine learning framework that residualizes standard outcome-above-expectation metrics to support valid frequentist inference and player-specific effect estimation in sports analytics.
Introduces self-separated and self-connected missingness models for mediator and outcome missingness in mediation analysis, enabling identification via conditional independences or shadow variables and extending shadow variable theory.
AC-IHT is a two-stage iterative algorithm for contaminated high-dimensional regression that attains minimax near-optimal rates, signal adaptivity under suitable conditions, and the strong oracle property.
citing papers explorer
-
The Statistical Cost of Adaptation in Multi-Source Transfer Learning
Multi-source transfer learning incurs an intrinsic adaptation cost that can exceed one, with phase transitions separating regimes where bias-agnostic estimators match oracle performance from those where they cannot.
-
Statistically and Computationally Optimal Estimation and Inference of Common Subspaces
Establishes statistical and computational optimality thresholds for common subspace estimation and inference under varying SNR regimes, including an impossibility result for adaptive confidence intervals below strong inference SNR.
-
Nonparametric Inference for Balance in Signed Networks
Introduces a nonparametric inference procedure based on a sparse signed graphon model that yields valid confidence intervals for balance parameters and reports strong empirical evidence for balance theory across real signed networks.
-
Bayesian Model Averaging under Predictor Redundancy via Density-Ratio Posterior Compression
A density-ratio framework compresses BMA posteriors into hard or soft support regions with explicit TV, KL, and predictive distortion bounds under predictor redundancy.
-
Causal Forecasting in Panel Data: A Two-Way Synthetic Forecasting Approach
Develops TWSF estimator for causal forecasting in panel data by combining synthetic controls with time-series models under low-rank latent factor assumptions, providing finite-sample bounds and asymptotic normality.
-
Two-Sample Hypothesis Testing for Subspace Equality in Network Data
A two-sample test for subspace equality in networks uses the Frobenius norm of projection matrix differences, with proven asymptotic normality to Gaussian under logarithmic average degree growth.
-
On Asymptotic Outlier Rejection in Bayesian Mixed Poisson Regression Models Under Extreme Target and Covariate Values
Mixed Poisson regression models with Gaussian latent variables are asymptotically robust to infinite target values but not to infinite covariate values, as shown for Poisson-Gamma, Poisson-log-t, and Poisson-RSB targets.
-
AI4BayesCode: From Natural Language Descriptions to Validated Modular Stateful Bayesian Samplers
AI4BayesCode generates validated modular stateful MCMC samplers from natural language Bayesian model descriptions via LLM translation, modular blocks, and recursive stateful composition.
-
Robust Learning of Heterogeneous Dynamic Systems
A distributionally robust ODE learning framework for heterogeneous systems that uses worst-case optimization over convex derivative combinations to produce a stabilized weighted estimator with theoretical guarantees.
-
StatsClaw: An AI-Collaborative Workflow for Statistical Software Development
StatsClaw is a multi-agent AI workflow that separates code implementation, data simulation, and testing via information barriers to produce faithful statistical software while keeping human control over methodological choices.
-
Rethinking player evaluation in sports: Goals above expectation and beyond
A double machine learning framework that residualizes standard outcome-above-expectation metrics to support valid frequentist inference and player-specific effect estimation in sports analytics.
-
Self-separated and self-connected models for mediator and outcome missingness in mediation analysis
Introduces self-separated and self-connected missingness models for mediator and outcome missingness in mediation analysis, enabling identification via conditional independences or shadow variables and extending shadow variable theory.
-
Adversarial Contamination Meets Hard Thresholding: An Iterative Algorithm with Signal Adaptivity and Minimax Optimality
AC-IHT is a two-stage iterative algorithm for contaminated high-dimensional regression that attains minimax near-optimal rates, signal adaptivity under suitable conditions, and the strong oracle property.