The Annals of Statistics 2024

31 papers
The 2024 edition of The Annals of Statistics comprises 31 papers focusing on theoretical advancements in high-dimensional inference and reinforcement learning. A significant portion of the research addresses offline reinforcement learning, exploring topics such as distribution correction, sample-optimal learning, and pessimism for robustness. Other prominent themes include matrix and tensor estimation, specifically regarding spectral bias correction in high-dimensional covariance and inference for heteroskedastic PCA with missing data. Methodological developments published this year cover asymptotic normality for statistical inference and higher-order asymptotics for regularized estimation. Specific studies investigate convergence rates for oblique regression trees and rank-based indices for testing independence between high-dimensional vectors. Notable contributors to this volume include Yanjun Han, Jianqing Fan, Yuxin Chen, and Aaditya Ramdas. The research also examines statistical-computational trade-offs in tensor PCA and the application of locally private estimation, providing new insights into the complexity measures required for modern statistical challenges.

Top Research Directions

The 35 research directions with The Annals of Statistics 2024 papers, ranked by paper count, next to 2023. Share is the fraction of the edition’s papers in the direction.

#Direction20242023Share 2024Share 20232023 rank
1Proportional Limit Scaling in High Dimensional Learning206.5%0.0%—
2The Statistical Speed Limits of Machine Learning206.5%0.0%—
3Bridging the Tabular Gap in Reinforcement Learning Theory103.2%0.0%—
4Causal Discovery with Feedback Loops103.2%0.0%—
5Causal Inference with Ancestral Graph Representations113.2%3.3%9
6Completeness in Causal Identification103.2%0.0%—
7Concentration Inequalities for High Dimensional Systems103.2%0.0%—
8Distribution Shift in Offline Reinforcement Learning103.2%0.0%—
9Fundamental Limits of the Privacy Utility Tradeoff103.2%0.0%—
10Generalizing Matrix and Tensor Recovery to Asymmetric Systems103.2%0.0%—
11Generalizing Neural Network Guarantees Beyond ReLU103.2%0.0%—
12Hybrid Linear Quadratic Loss Functions for Robust Optimization103.2%0.0%—
13Intrinsic Dimension Adaptation in Deep Learning103.2%0.0%—
14Mathematical Foundations of Transfer Learning Generalization103.2%0.0%—
15Matrix Recovery with Approximately Low Rank Data103.2%0.0%—
16Metric Differential Privacy103.2%0.0%—
17Mitigating Combinatorial Explosion in Causal Discovery103.2%0.0%—
18Non Asymptotic Error Bounds for Finite Datasets103.2%0.0%—
19Nonparametric Estimation under Unknown Function Smoothness113.2%3.3%31
20Oblique Hyperplane Splits for Decision Trees103.2%0.0%—
21Offline Reinforcement Learning under Single Policy Coverage103.2%0.0%—
22Optimization and Generalization with Non-Smooth Activations103.2%0.0%—
23Pessimistic Value Estimation for Offline Reinforcement Learning103.2%0.0%—
24Proving Convergence Under Iterative Data Reuse103.2%0.0%—
25Proving Streaming Space Lower Bounds via Communication Complexity103.2%0.0%—

Download all directions as CSV

Institutions

Institutions credited by The Annals of Statistics 2024 papers: a paper credits every institution named in its affiliations, once. 32 institutions share 22 credited papers.

#InstitutionPapersShare20232023 rank
1Stanford University 522.7%52
2University of Pennsylvania 522.7%43
3Princeton University 313.6%26
4University of Southern California 313.6%135
5Carnegie Mellon University 29.1%145
6Courant Institute of Mathematical Sciences 29.1%0—
7Massachusetts Institute of Technology 29.1%34
8New York University 29.1%0—
9University of California, Berkeley 29.1%61
1002xkdev90 14.5%0—
11Adobe Systems (United States) 14.5%0—
12Australian National University 14.5%0—
13California Institute of Technology 14.5%0—
14Carnegie Mellon University Qatar 14.5%0—
15Chinese University of Hong Kong 14.5%119
16Columbia University 14.5%115
17Deutsches Zentrum für Luft- und Raumfahrt e. V. (DLR) 14.5%0—
18Meta (United States) 14.5%126
19Michigan State University 14.5%0—
20Microsoft (United States) 14.5%0—
21Peking University 14.5%0—
22Rutgers, The State University of New Jersey 14.5%144
23Tsinghua University 14.5%0—
24University of California San Diego 14.5%0—
25University of California, Davis 14.5%143
26University of Chicago 14.5%27
27University of Edinburgh 14.5%125
28University of Michigan 14.5%116
29University of Warwick 14.5%0—
30University of Waterloo 14.5%123

Download all institutions as CSV

About this data

Papers. The Annals of Statistics 2024 lists 31 papers on Lacuna, duplicate versions counted once; rejected and withdrawn submissions are left out.

Directions. Directions are Lacuna’s clustering of the concept elements extracted from each paper, so a paper can belong to several and direction counts add up to more than the paper count. A direction’s share is its papers divided by the edition’s papers. Comparisons with 2023 use shares, since the edition went from 30 to 31 papers.

Institutions. A paper credits each institution named in its affiliations, resolved to a ROR record, once. 22 of 31 papers (71.0%) credit at least one institution and are the base for institution shares; the rest have no resolved affiliation yet.

Data. Counts come from Lacuna’s data build of 2026-11-21 and change as papers are added. Tables as CSV: directions · institutions. Cite as: Lacuna, “The Annals of Statistics 2024”, https://lacuna.tiptreesystems.com/venue/the-annals-of-statistics/2024.

Top Authors

AuthorPapers
Aaditya Ramdas2
Jianqing Fan2
Yuxin Chen2
Andreas Gerhardus1
Bharath Sriperumbudur1
Changxiao Cai1
Daniel Hsu1
Hamed Hassani1
Hans-Georg Müller1
Jason Matthew Klusowski1
Jiantao Jiao1
Kannan Ramchandran1
Timothy Ivor Cannings1
Youssef Mroueh1
Yuejie Chi1
Yuting Wei1
Zhengxin Zhang1
Ziv Goldfeld1

Impactful Papers

Papers from The Annals of Statistics 2024 that shaped research clusters.

Influential

Papers

Title
A new test for high-dimensional two-sample mean problems with consideration of correlation structure
Characterization of causal ancestral graphs for time series with latent confounders
Computational lower bounds for graphon estimation via low-degree polynomials
Convergence rates of oblique regression trees for flexible function libraries
Correction to “Nonparametric regression using deep neural networks with ReLU activation function”
Deep nonlinear sufficient dimension reduction
Dimension free ridge regression
Efficient and multiply robust risk estimation under general forms of dataset shift
Fundamental limits of low-rank matrix estimation with diverging aspect ratios
Gromov-Wasserstein Distances: Entropic Regularization, Duality, and Sample Complexity
How do noise tails impact on deep ReLU networks?
Increasing dimension asymptotics for two-way crossed mixed effect models
Inference for heteroskedastic PCA with missing data
ℓ2 inference for change points in high-dimensional time series via a Two-Way MOSUM
Metric statistics: Exploration and inference for random objects with distance profiles
Noisy recovery from random linear observations: Sharp minimax rates under elliptical constraints
Nonparametric classification with missing data
On the existence of powerful p-values and e-values for composite hypotheses
On the statistical complexity of sample amplification
Rank-based indices for testing independence between two high-dimensional vectors
Settling the sample complexity of model-based offline reinforcement learning
Statistical complexity and optimal algorithms for nonlinear ridge bandits
Statistical-computational trade-offs in tensor PCA and related problems via communication complexity
Statistical inference for decentralized federated learning
Statistical inference for four-regime segmented regression models
Testing high-dimensional regression coefficients in linear models
The curse of overparametrization in adversarial training: Precise analysis of robust generalization for random features regression
The right complexity measure in locally private estimation: It is not the Fisher information
Time-uniform central limit theory and asymptotic confidence sequences
Transfer learning for contextual multi-armed bandits
Transfer learning for functional mean estimation: Phase transition and adaptive algorithms