Search:
Match:
4 results
Research Paper#Machine Learning, High-Dimensional Statistics, Sparse Learning, Weak Supervision🔬 ResearchAnalyzed: Jan 3, 2026 17:12

Sparse Classification with Positive-Confidence Data in High Dimensions

Published:Dec 30, 2025 19:53
1 min read
ArXiv

Analysis

This paper addresses the challenge of high-dimensional classification when only positive samples with confidence scores are available (Positive-Confidence or Pconf learning). It proposes a novel sparse-penalization framework using Lasso, SCAD, and MCP penalties to improve prediction and variable selection in this weak-supervision setting. The paper provides theoretical guarantees and an efficient algorithm, demonstrating performance comparable to fully supervised methods.
Reference

“The paper proposes a novel sparse-penalization framework for high-dimensional Pconf classification.”

PermalinkArXiv
Research Paper#Spatial Statistics, Banking, Machine Learning🔬 ResearchAnalyzed: Jan 3, 2026 19:01

Bank Location Modeling with Sparse Group Lasso

Published:Dec 29, 2025 08:26
1 min read
ArXiv

Analysis

This paper applies a statistical method (sparse group Lasso) to model the spatial distribution of bank locations in France, differentiating between lucrative and cooperative banks. It uses socio-economic data to explain the observed patterns, providing insights into the banking sector and potentially validating theories of institutional isomorphism. The use of web scraping for data collection and the focus on non-parametric and parametric methods for intensity estimation are noteworthy.
Reference

“The paper highlights a clustering effect in bank locations, especially at small scales, and uses socio-economic data to model the intensity function.”

PermalinkArXiv
Research Paper#Compressed Sensing, Sparse Recovery, Optimization, Image Reconstruction🔬 ResearchAnalyzed: Jan 3, 2026 19:10

DCEN for Compressed Sensing

Published:Dec 29, 2025 01:35
1 min read
ArXiv

Analysis

This paper introduces a novel framework, DCEN, for sparse recovery, particularly beneficial for high-dimensional variable selection with correlated features. It unifies existing models, provides theoretical guarantees for recovery, and offers efficient algorithms. The extension to image reconstruction (DCEN-TV) further enhances its applicability. The consistent outperformance over existing methods in various experiments highlights its significance.
Reference

“DCEN consistently outperforms state-of-the-art methods in sparse signal recovery, high-dimensional variable selection under strong collinearity, and Magnetic Resonance Imaging (MRI) image reconstruction, achieving superior recovery accuracy and robustness.”

PermalinkArXiv
Research#Bayesian Lasso🔬 ResearchAnalyzed: Jan 10, 2026 08:18

Analyzing Convergence in Bayesian Lasso with Data Augmentation

Published:Dec 23, 2025 04:18
1 min read
ArXiv

Analysis

This research focuses on the theoretical underpinnings of data augmentation techniques within a specific statistical modeling context. The study of convergence is crucial for establishing the reliability and efficiency of these methods.
Reference

“The article is from ArXiv, indicating a pre-print publication likely targeting a specialized audience.”

PermalinkArXiv