Model-free Multiple Testing for Matrix-valued Predictors with False Discovery Control

A methodological paper and associated materials proposing a novel procedure for variable selection in high-dimensional matrix-valued data. The work, authored by Lei Yan and last updated in May 2026, includes a PDF document, markdown files, and a ZIP archive totaling 14.6 MB. The method is designed to control the false discovery rate for simultaneous row and column selection without requiring p-value computation.

Use Cases

• Developing new multiple testing procedures based on the described folding selection subspaces (FSS) framework.
• Benchmarking false discovery rate control methods for matrix predictor regression against the proposed approach.
• Applying the model-free selection technique to real-world datasets with matrix-valued predictors.

Strengths

• The methodology is described in a 14.6 MB package containing a PDF, markdown, and ZIP files.
• The proposed procedure is model-free and designed to bypass p-value computation.
• Theoretical contributions include proofs of asymptotic distribution and FDR control.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset appears to be a research paper and its supplementary materials rather than a primary data table.

Provenance

Source

Lei Yan via figshare

Collection Method

Likely generated for the associated research publication.

Time Range

null

Freshness

Last updated 2026-05-08 16:27:51; freshness should be verified.

Geography

null