Distributional Split Criteria for Random Forests: Extensions, Shrinkage, and the Robustness of Mean Splitting

• URL: http://arxiv.org/abs/2607.23721v1
• Date: Sun, 26 Jul 2026 15:37:30 GMT
• Title: Distributional Split Criteria for Random Forests: Extensions, Shrinkage, and the Robustness of Mean Splitting
• Abstract summary: Distributional random forests replace mean-based CART splitting with criteria that compare the full conditional response distribution in candidate children.<n>We implement and systematically study a family of such criteria inside a single honest-forest implementation.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Distributional random forests replace mean-based CART splitting with criteria that compare the full conditional response distribution in candidate children. We implement and systematically study a family of such criteria inside a single honest-forest implementation: isotropic random-Fourier-feature maximum mean discrepancy (MMD), an anisotropic diagonal-bandwidth variant, an adaptive per-split frequency-selection variant, and a non-kernel sliced-Wasserstein criterion, together with post-hoc kernel-mean shrinkage of the forest weights. Using paired-seed comparisons across synthetic quantile mechanisms, real univariate benchmarks, a California-housing subsample curve, and multivariate synthetic and real responses, we characterize where each extension pays. Three findings recur. First, among distributional criteria ordinary isotropic MMD is already close to best in class: the anisotropic, adaptive-frequency, and sliced-Wasserstein extensions, and post-hoc shrinkage, do not systematically improve on it. Second, on scalar tabular regression mean-based CART splitting remains the robust default and wins many cells. Third, multivariate responses are the regime where distributional splitting clearly earns its keep, most sharply on a pure-dependence copula where the energy score separates the criteria even though marginal CRPS does not. The evidence supports a simple allocation story: distributional splitting helps only when non-location structure is both present and estimable; otherwise it dilutes split-selection power away from the mean. All criteria, the honest forest, and the paired-comparison harness are implemented in the open-source \texttt{drforest} library, whose Rust-backed split search makes broad criterion sweeps inexpensive.

Related papers

• When to Truncate a Feature Ranking: A Residual-Overlap Stopping Rule for Subset Selection [0.0]
  This paper develops a distributional framework for transforming supervised feature rankings into class-independent subsets.<n>An empirical comparison on high-dimensional genomic datasets illustrates that the rule can reduce tens of thousands of variables to a few dozen.
  arXiv  Detail & Related papers  (2026-06-30T14:00:29Z)
• CriterAlign: Criterion-Centric Rationale Alignment for Code Preference Judging [95.02210956333374]
  We propose a criterion-centric framework that adapts rubric-based judging to pairwise preference evaluation.<n>On BigCodeReward, CriterAlign improves a Qwen2.5-VL-32B monolithic judge from 60.4% to 66.3% accuracy.
  arXiv  Detail & Related papers  (2026-05-19T10:59:19Z)
• DISA: Offline Importance Sampling for Distribution-Matching LLM-RL [56.9445657766829]
  We introduce DISA, which moves this calibration problem outside the RL loop.<n> DISA draws proposal trajectories offline, estimates the partition function via importance sampling, and freezes the resulting partition-function estimate.<n>On two open-weight backbones across six math and three code benchmarks, DISA matches or exceeds the online-coupled distribution-matching baseline Flow.
  arXiv  Detail & Related papers  (2026-05-17T07:14:44Z)
• Closed-Form Beta Distribution Estimation from Sparse Statistics with Random Forest Implicit Regularization [0.0]
  This work advances distribution recovery from sparse data and ensemble classification through three main contributions.<n>First, we introduce a closed-form estimator that reconstructs scaled beta distributions from limited statistics.<n>Second, we establish a link between classification accuracy and distributional closeness by deriving error bounds.<n>Third, we show that zero-variance features act as an implicit regularizer, increasing selection probability for mid-ranked predictors.
  arXiv  Detail & Related papers  (2025-07-31T17:55:07Z)
• ReCAB-VAE: Gumbel-Softmax Variational Inference Based on Analytic Divergence [17.665255113864795]
  We present a novel divergence-like metric which corresponds to the upper bound of the Kullback-Leibler divergence (KLD) of a relaxed categorical distribution. We also propose a relaxed categorical analytic bound variational autoencoder (ReCAB-VAE) that successfully models both continuous and relaxed latent representations.
  arXiv  Detail & Related papers  (2022-05-09T08:11:46Z)
• Random Forest Weighted Local Fréchet Regression with Random Objects [18.128663071848923]
  We propose a novel random forest weighted local Fr'echet regression paradigm.<n>Our first method uses these weights as the local average to solve the conditional Fr'echet mean.<n>Second method performs local linear Fr'echet regression, both significantly improving existing Fr'echet regression methods.
  arXiv  Detail & Related papers  (2022-02-10T09:10:59Z)
• Exact Recovery in the General Hypergraph Stochastic Block Model [92.28929858529679]
  This paper investigates fundamental limits of exact recovery in the general d-uniform hypergraph block model (d-HSBM) We show that there exists a sharp threshold such that exact recovery is achievable above the threshold and impossible below it.
  arXiv  Detail & Related papers  (2021-05-11T03:39:08Z)
• A Unified Joint Maximum Mean Discrepancy for Domain Adaptation [73.44809425486767]
  This paper theoretically derives a unified form of JMMD that is easy to optimize. From the revealed unified JMMD, we illustrate that JMMD degrades the feature-label dependence that benefits to classification. We propose a novel MMD matrix to promote the dependence, and devise a novel label kernel that is robust to label distribution shift.
  arXiv  Detail & Related papers  (2021-01-25T09:46:14Z)
• Consensus-Guided Correspondence Denoising [67.35345850146393]
  We propose to denoise correspondences with a local-to-global consensus learning framework to robustly identify correspondence. A novel "pruning" block is introduced to distill reliable candidates from initial matches according to their consensus scores estimated by dynamic graphs from local to global regions. Our method outperforms state-of-the-arts on robust line fitting, wide-baseline image matching and image localization benchmarks by noticeable margins.
  arXiv  Detail & Related papers  (2021-01-03T09:10:00Z)
• Distributional Random Forests: Heterogeneity Adjustment and Multivariate Distributional Regression [0.8574682463936005]
  We propose a novel forest construction for multivariate responses based on their joint conditional distribution. The code is available as Python and R packages drf.
  arXiv  Detail & Related papers  (2020-05-29T09:05:00Z)
• Convolution-Weight-Distribution Assumption: Rethinking the Criteria of Channel Pruning [90.2947802490534]
  We find two blind spots in the study of pruning criteria. The ranks of filters'Importance Score are almost identical, resulting in similar pruned structures. The filters'Importance Score measured by some pruning criteria are too close to distinguish the network redundancy well.
  arXiv  Detail & Related papers  (2020-04-24T09:54:21Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.