Research map / Bandits and online learning / Online convex optimization and games
Online Sufficient Dimension Reduction Through Sliced Inverse Regression
Zhanrui Cai, Runze Li, Liping Zhu
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 |
|---|---|---|---|
| use online gradient descent SVD instead of perturbation | Online SIR · Simulation model (14), t=10000 | distance d(B,B_hat): 0.3638 -> 0.0479 | Improved |
| use online gradient descent SVD instead of perturbation | Online SIR · Simulation model (12), t=10000 | distance d(B,B_hat): 0.0035 -> 0.0112 | Hurt |
| use online gradient descent SVD instead of perturbation | Online SIR · Simulation model (12), t=10000 | computation time (s): 59.2208 -> 1.5360 | Improved |
| use online SIR (perturbation) instead of batch SIR rerun per observation | Online SIR · Simulation model (12), t=10000 | computation time (s): 1204.12 -> 59.2208 | Improved |
| use online SIR instead of block-wise SIR | Online SIR perturbation · Simulation model (12), t=10000 | distance d(B,B_hat): 0.1431 -> 0.0035 | Improved |
| use supervised SIR instead of online PCA | Online SIR perturbation · Simulation model (12), t=10000 | distance d(B,B_hat): 0.8521 -> 0.0035 | Improved |
| increase sample size from 1000 to 10000 | Online SIR perturbation · Simulation model (12) | distance d(B,B_hat): 0.0276 -> 0.0035 | Improved |
| increase sample size from 1000 to 10000 | Block-wise SIR · Simulation model (12) | distance d(B,B_hat): 0.2256 -> 0.1431 | No clear effect |
| use online SIR instead of SVM on all covariates | Online SIR gradient descent · housing (SVM prediction) | prediction error: 0.3373 vs 0.1867 | Mixed |
| apply dimension reduction with online SIR before SVM | Online SIR gradient descent · ozone (SVM prediction) | prediction error: 0.7468 -> 0.4844 | Improved |
| use online SIR instead of SVM on all covariates | Online SIR perturbation · digits (SVM prediction) | misclassification error: 0.0208 -> 0.1126 | Hurt |
| change classifier from SVM to random forest | Online SIR perturbation · digits | prediction error: 0.1126 -> 0.1069 | No clear effect |
Closest papers by meaning
- Online Directional Regression for Streaming Sufficient Dimension Reduction (NeurIPS 2026 · similarity 0.90)
- Abundant Inverse Regression Using Sufficient Reduction and Its Applications (ECCV 2016 · similarity 0.85)
- A Forward Approach for Sufficient Dimension Reduction in Binary Classification (JMLR 2022 · similarity 0.85)
- High-Dimensional Analysis for Generalized Nonlinear Regression: From Asymptotics to Algorithm (AAAI 2024 · similarity 0.84)
- Characterizing Overfitting in Kernel Ridgeless Regression Through the Eigenspectrum (ICML 2024 · similarity 0.83)
- Solving Sparse \& High-Dimensional-Output Regression via Compression (NeurIPS 2024 · similarity 0.83)
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.
