Treatment changes identify causal effects under two non-nested structural models that difference out time-constant confounders; under random walk on treatment these are equivalent to levels-based methods, and two-way fixed effects regression is doubly robust.
Generalized random forests.Ann
11 Pith papers cite this work, alongside 16 external citations. Polarity classification is still indexing.
11
Pith papers citing it
16
external citations · OpenAlex
citation-role summary
background 1
citation-polarity summary
years
2026 11roles
background 1polarities
background 1representative citing papers
CART random forests are analyzed as controlled stochastic processes, separating subsampling and split policy effects, with explicit MSE derivations for linear models showing local stabilization but potential global suboptimality.
Heat-kernel smoothing over weighted points on a compact manifold yields a scale-dependent geometric effective sample size that discounts nearby and duplicate particles.
Develops HDA and VDA methods plus adapted crossing-point asymptotics for locating and testing treatment-effect discontinuities in distributional treatment effects, illustrated on synthetic data and Mexico's PROGRESA program.
The paper develops set-valued policies and conformal policy learning methods that output treatment sets with marginal coverage guarantees for robust decision-making under uncertainty.
A placebo-anchored cross-fitted doubly robust estimator for heterogeneous treatment effects in meta-analysis under covariate shift that improves accuracy at small target sample sizes.
A penalized likelihood estimator for GEV parameters, weighted by generalized random forest weights, is introduced for extreme quantile regression to improve tail extrapolation and handle many predictors.
Backdoor-adjusted ATEs on 21,098 UK Biobank participants showed total femur BMC and BMD with the largest hip fracture risk reductions (-0.0047 per SD), and adding the top 11 phenotypes to clinical variables raised AUC to 0.842 versus FRAX 0.709.
Contract Scoring applies adaptive nearest neighbors on ensemble trees to grade enterprise contracts by historical peers, yielding letter grades and reported revenue gains at Databricks.
ADHD status carries a direct penalty of 0.67 points on high school STEM GPA, with 63% of the total disparity not explained by observed sociodemographic or academic factors.
Conditional inference forests rank competitively as top-k feature selectors in classification and regression benchmarks, with runtime factors identified but limited impact on scores.
citing papers explorer
-
When Do Treatment Changes Identify Causal Effects?
Treatment changes identify causal effects under two non-nested structural models that difference out time-constant confounders; under random walk on treatment these are equivalent to levels-based methods, and two-way fixed effects regression is doubly robust.
-
CART Random Forests as Sequential Allocation over Random Opportunity Sets: A Stochastic-Control Theory of Ensemble Risk
CART random forests are analyzed as controlled stochastic processes, separating subsampling and split policy effects, with explicit MSE derivations for linear models showing local stabilization but potential global suboptimality.
-
Heat-Kernel Entropy Profiles and Geometric Effective Sample Size for Weighted Measures on Manifolds
Heat-kernel smoothing over weighted points on a compact manifold yields a scale-dependent geometric effective sample size that discounts nearby and duplicate particles.
-
A Toolkit for the Study of Treatment-Effect Discontinuities
Develops HDA and VDA methods plus adapted crossing-point asymptotics for locating and testing treatment-effect discontinuities in distributional treatment effects, illustrated on synthetic data and Mexico's PROGRESA program.
-
Set-Valued Policy Learning
The paper develops set-valued policies and conformal policy learning methods that output treatment sets with marginal coverage guarantees for robust decision-making under uncertainty.
-
Transfer Learning for Meta-analysis Under Covariate Shift
A placebo-anchored cross-fitted doubly robust estimator for heterogeneous treatment effects in meta-analysis under covariate shift that improves accuracy at small target sample sizes.
-
Penalized estimation of GEV parameters for extreme quantile regression
A penalized likelihood estimator for GEV parameters, weighted by generalized random forest weights, is introduced for extreme quantile regression to improve tail extrapolation and handle many predictors.
-
DXA-Derived Skeletal Phenotypes and Hip Fracture Risk: A Backdoor-Adjusted Causal Analysis
Backdoor-adjusted ATEs on 21,098 UK Biobank participants showed total femur BMC and BMD with the largest hip fracture risk reductions (-0.0047 per SD), and adding the top 11 phenotypes to clinical variables raised AUC to 0.842 versus FRAX 0.709.
-
Algorithmic Contract Design at Scale: Adaptive Peer Comparison for Enterprise Pricing
Contract Scoring applies adaptive nearest neighbors on ensemble trees to grade enterprise contracts by historical peers, yielding letter grades and reported revenue gains at Databricks.
-
Causal Fairness Analysis of ADHD Status and High School STEM Outcomes
ADHD status carries a direct penalty of 0.67 points on high school STEM GPA, with 63% of the total disparity not explained by observed sociodemographic or academic factors.
-
Conditional Inference Trees and Forests for Feature Selection
Conditional inference forests rank competitively as top-k feature selectors in classification and regression benchmarks, with runtime factors identified but limited impact on scores.