Research map / Learning theory and algorithms / Kernels and regression theory
A Regularization-Based Adaptive Test for High-Dimensional GLMs
Chong Wu, Gongjun Xu, Xiaotong Shen, Wei Pan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| estimate nuisance parameters with TLP penalty instead of full MLE (ablation) | aiSPU · G x E, n=2000, p=300 | Type I error rate: 0.551 -> 0.042 | Improved |
| estimate nuisance parameters with TLP penalty instead of full MLE (ablation) | aiSPU · G x E, n=2000, p=300, q1=q2=30 | Type I error rate: 0.712 -> 0.057 | Improved |
| use TLP instead of Lasso or ridge penalty (ablation) | aiSPU · G x E, n=2000, p=300, q1=q2=30 | Type I error rate / power: 0.016 -> 0.057 | Improved |
| use TLP instead of MCP or SCAD penalty (ablation) | aiSPU · G x E simulations | power | Improved |
| use TLP instead of SCAD penalty (ablation) | aiSPU · linear model simulations | Type I error rate | Improved |
| use simulation-based p-values instead of theoretical null for GESAT | GESAT · G x E, n=2000, p=300 | Type I error rate: 0.636 -> 0.044 | Improved |
| adapt over gamma (aiSPU) instead of fixed iSPU(gamma) | iSPU · G x E, n=200, p=1000, s=0.005, c=2 | power (%): iSPU(inf) 85 vs aiSPU 79.4 | Mixed |
| use gamma = infinity for very sparse signals | iSPU · G x E, n=200, p=1000, s=0.001 and 0.005 | power | Improved |
| use gamma = 2 for dense signals | iSPU · G x E, s=0.2 | power | Improved |
| use aiSPU instead of GESAT on real ADNI1 pathway | aiSPU · ADNI1 hsa00051 | p-value: 0.016 -> 0.0003 | Improved |
| violate sparse-effect assumption with many small effects in Z | aiSPU · G x E simulations (omnigenic) | Type I error / power | No clear effect |
| use UminP instead of aiSPU when signals are dense | UminP · G x E, s=0.05 and 0.2 | power | Hurt |
Closest papers by meaning
- Optimal Minimax Variable Selection for Large-Scale Matrix Linear Regression Model (JMLR 2021 · similarity 0.86)
- Modelling Interactions in High-dimensional Data with Backtracking (JMLR 2016 · similarity 0.85)
- Graphical Lasso and Thresholding: Equivalence and Closed-form Solutions (JMLR 2019 · similarity 0.85)
- Safe Feature Screening for Generalized LASSO (TPAMI 2017 · similarity 0.85)
- The Complete Lasso Tradeoff Diagram (NeurIPS 2020 · similarity 0.84)
- Testing for Differences in Gaussian Graphical Models: Applications to Brain Connectivity (NeurIPS 2016 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
