Feature selection when there are many influential features

DOI10.3150/13-BEJ536zbMath1398.62162arXiv0911.4076MaRDI QIDQ396025

Hugh Miller, Jiashun Jin, Hall, Peter

Publication date: 8 August 2014

Published in: Bernoulli (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/0911.4076

zbMATH Keywords

classificationrankingdimension reductionmaximum likelihoodthresholdingfeature selectionlogit modelchange-point analysis

Mathematics Subject Classification ID

Nonparametric hypothesis testing (62G10) Classification and discrimination; cluster analysis (statistical aspects) (62H30) Statistical ranking and selection procedures (62F07)

Related Items

Detection boundary and higher criticism approach for rare and weak genetic effects, Linear Hypothesis Testing in Dense High-Dimensional Linear Models

Uses Software

• PDCO
• ElemStatLearn
• boost

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Sure independence screening in generalized linear models with NP-dimensionality
• Strong approximations of level exceedences related to multiple hypothesis testing
• High-dimensional classification using features annealed independence rules
• Heuristics of instability and stabilization in model selection
• Some theory for Fisher's linear discriminant function, `naive Bayes', and some alternatives when there are many more variables than observations
• Inference for change point and post change means after a CUSUM test.
• The Dantzig selector: statistical estimation when \(p\) is much larger than \(n\). (With discussions and rejoinder).
• Adapting to unknown sparsity by controlling the false discovery rate
• Higher criticism thresholding: Optimal feature selection when useful features are rare and weak
• Impossibility of successful classification when useful features are rare and weak
• Better Subset Regression Using the Nonnegative Garrote
• Recovery of Short, Complex Linear Combinations Via<tex>$ell _1$</tex>Minimization
• Feature selection by higher criticism thresholding achieves the optimal phase diagram
• Atomic Decomposition by Basis Pursuit
• Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties
• Uncertainty principles and ideal atomic decomposition
• Optimally sparse representation in general (nonorthogonal) dictionaries via ℓ ¹ minimization
• For most large underdetermined systems of linear equations the minimal 𝓁₁‐norm solution is also the sparsest solution
• For most large underdetermined systems of equations, the minimal 𝓁₁‐norm near‐solution approximates the sparsest near‐solution
• The elements of statistical learning. Data mining, inference, and prediction