@StatMLPapers

Unofficial updates of statistical machine learning papers on arXiv

arxiv.org/list/stat.ML/recent
Joined June 2012
  • Tweets72.8K
  • Following0
  • Followers25.7K
Extreme classification: beating chance with one training example from each class ift.tt/VmBYi8H

Extreme classification: beating chance with one training example...

We study a minimal classification problem: Given independent labeled observations $X\sim P$ and $Z\sim Q$ from two unknown distributions $P,Q$, and given an independent target $Y$ drawn with equal...

arxiv.org
2
1
14
1,663
Complex Problem Solving in Large Language Models: A Statistical Control Survey and Diagnostic Framework ift.tt/RgwJBND

Complex Problem Solving in Large Language Models: A Statistical...

Complex problem solving (CPS) with large language models (LLMs) is often framed as a matter of stronger reasoning or longer generation. Yet early-step error amplification, prompt brittleness, and...

arxiv.org
2
3
1
29
2,130
A Smoothed Discrepancy Principle for Random Feature Methods and Neural Networks ift.tt/Qo9du7Z

A Smoothed Discrepancy Principle for Random Feature Methods and...

We study data-driven early stopping for spectral regularisation methods in the classical non-parametric regression setting. Building on the discrepancy principle, we propose a multi-scale stopping...

arxiv.org
1,187
Triply-Scalable Equivariant Gaussian Process Modeling ift.tt/B4rC0L7

Triply-Scalable Equivariant Gaussian Process Modeling

Gaussian processes (GPs) provide principled probabilistic predictions while encoding prior knowledge, including equivariances. Yet, their use in large-scale scientific problems is limited by...

arxiv.org
1
1,029
Diagonalized Attention for Individualized Regression: Latent-Row Localization and Prediction ift.tt/6jImPTG

Diagonalized Attention for Individualized Regression: Latent-Row...

Modern text and image representations are often matrix-valued, with rows corresponding to tokens, patches, or other local feature vectors. Predictive information is often sparse but...

arxiv.org
1,013
Sparse Identification for Automatic Large-Scale Screening: A Constraint-Aware Framework with Ultra Fast Decoding Algorithm ift.tt/hyI6EY2

Sparse Identification for Automatic Large-Scale Screening: A...

In the early stages of a pandemic, identification of a small number of infected individuals through large-scale screening is critical for pandemic control, yet remains challenging under limited...

arxiv.org
3
996
Brownian Heads for Deep ReLU Representations: Activation Mass and the Cost of Same-Sample Selection ift.tt/nSjZDL0

Brownian Heads for Deep ReLU Representations: Activation Mass and...

Deep representation learning often selects hidden features and fits the final predictor on the same sample, so fixed-feature analysis performed after selection can omit selection cost. We study...

arxiv.org
996
Improving the Predictive Performance of Bootstrap Aggregating by Dirichlet Resampling ift.tt/J1OitdT

Improving the Predictive Performance of Bootstrap Aggregating by...

We revisit Breiman's observation that reducing inter-tree correlation without weakening individual trees can improve random forests. Building on this principle, we introduce two variants:...

arxiv.org
939
How Many Posterior Samples? Calibrated Stopping for Adaptive Sensing ift.tt/fCdyE2Y

How Many Posterior Samples? Calibrated Stopping for Adaptive Sensing

In classification-oriented adaptive sensing, posterior samples characterize uncertainty at the current measurement state and can serve two roles: they may guide the next sensing direction, while...

arxiv.org
946
Riemannian Simultaneous Inference for Tangent Vector Field Regression ift.tt/KuAornS

Riemannian Simultaneous Inference for Tangent Vector Field Regression

We consider nonparametric tangent vector field regression on a Riemannian manifold without boundary. Because responses at different points lie in different tangent spaces, the proposed kernel...

arxiv.org
11
1,200
Training-Adaptive Convolutional Sparse Coding via Information Bottleneck for Robust Visual Representation ift.tt/T9hgpSK

Training-Adaptive Convolutional Sparse Coding via Information...

Visual signals require compact yet sufficient representations for robust downstream prediction. Convolutional sparse coding (CSC) provides an explicit mechanism for suppressing redundant...

arxiv.org
951
Aggregated Posterior Predictive Checks for Generative Modeling ift.tt/OVL7S50

Aggregated Posterior Predictive Checks for Generative Modeling

Latent variable generative models are commonly fit using simple priors over latent variables, but draws from these priors often fail to produce realistic data. This failure is due to a mismatch...

arxiv.org
922
Layerwise Decoupling for Stable Structured Sparsification of Fully Connected Layers ift.tt/qCLdoUk

Layerwise Decoupling for Stable Structured Sparsification of Fully...

We propose a decoupled, layerwise method for structurally sparsifying the fully connected layers of pretrained neural networks. Rather than penalizing all layers jointly, our approach extracts...

arxiv.org
1
4
962
Repulsive normalizing flow mixtures for adaptive importance sampling: reliability analysis of complex systems ift.tt/gbiZCYX

Repulsive normalizing flow mixtures for adaptive importance...

Accurate rare-event estimation can be computationally expensive. Classical adaptive importance sampling (IS) schemes often rely on restrictive proposal families and can struggle under multiple...

arxiv.org
1
943
Robust Dual-Regularized Variable Selection under Outlier Contamination ift.tt/YNVL7UM

Robust Dual-Regularized Variable Selection under Outlier Contamination

Real data often contain unusual observations that can exert disproportionate effects on variable selection, especially in complex predictor settings. We propose a two-stage {\it sparse median...

arxiv.org
915
Multi-Domain Clustering via Measure Quantization ift.tt/N4kCPfi

Multi-Domain Clustering via Measure Quantization

Clustering is a fundamental task in data analysis, typically addressed through centroid-based methods such as K-means. In this work, we present a general framework for multi-domain clustering via...

arxiv.org
2
12
1,067
Single-Loop Stochastic Projected Damped Extragradient Methods for Stochastic Nonconvex--(Strongly) Concave Minimax Optimization ift.tt/ExnopSJ

Single-Loop Stochastic Projected Damped Extragradient Methods for...

We develop single-loop stochastic projected damped extragradient methods for nonconvex--(strongly) concave minimax optimization, with complexity guarantees for both game stationarity (GS) and...

arxiv.org
3
947
Neural composite likelihood estimation: simulation based inference for time series ift.tt/c6Jo3yM

Neural composite likelihood estimation: simulation based inference...

Simulation based inference (SBI) circumvents the challenge of intractable likelihoods by using a simulator that generates data given parameter values. For instance, neural likelihood estimation...

arxiv.org
925
EnterpriseVal: Quantifying the Efficacy, Reliability and Value of Generative AI in the Enterprise ift.tt/oUc10JK

EnterpriseVal: Quantifying the Efficacy, Reliability and Value of...

Frontier language models now produce professional deliverables that expert graders judge to match human work on a substantial share of economically valuable tasks, yet most enterprise GenAI...

arxiv.org
1
3
920
Geometric Mean Pooling for Equal-Weight Multiplicative Coarse-Graining ift.tt/qwmpsxh

Geometric Mean Pooling for Equal-Weight Multiplicative Coarse-Graining

As an alternative to the additive and extremal biases of average and max pooling, we introduce Geometric Mean Pooling (GMP), a signed pooling operator that combines the product of feature signs...

arxiv.org
1
9
986
Load more