The optimal ridge penalty for real-world high-dimensional data can be zero or negative due to the implicit ridge regularization
From MaRDI portal
(Redirected from Publication:5148955)
Recommendations
- Penalization-induced shrinking without rotation in high dimensional GLM regression: a cavity analysis
- High dimensional thresholded regression and shrinkage effect
- Combination of nonconvex penalties and ridge regression for high-dimensional linear models
- A small-sample choice of the tuning parameter in ridge regression
- Geometry and properties of generalized ridge regression in high dimensions
Cites work
- A jamming transition from under- to over-parametrization affects generalization in deep learning
- An introduction to statistical learning. With applications in R
- Atomic decomposition by basis pursuit
- High-dimensional asymptotics of prediction: ridge regression and classification
- High-dimensional dynamics of generalization error in neural networks
- scientific article; zbMATH DE number 6378127 (Why is no real title available?)
- scientific article; zbMATH DE number 845714 (Why is no real title available?)
- scientific article; zbMATH DE number 1391397 (Why is no real title available?)
- scientific article; zbMATH DE number 6438182 (Why is no real title available?)
- scientific article; zbMATH DE number 3227378 (Why is no real title available?)
- Just interpolate: kernel ``ridgeless regression can generalize
- Reconciling modern machine-learning practice and the classical bias-variance trade-off
- Ridge Regression: Biased Estimation for Nonorthogonal Problems
- The Dantzig selector: statistical estimation when \(p\) is much larger than \(n\). (With discussions and rejoinder).
- The implicit bias of gradient descent on separable data
Cited in
(23)- Double data piling leads to perfect classification
- Surprises in high-dimensional ridgeless least squares interpolation
- scientific article; zbMATH DE number 7370646 (Why is no real title available?)
- Generalization error rates in kernel regression: the crossover from the noiseless to noisy regime*
- Binary classification of Gaussian mixtures: abundance of support vectors, benign overfitting, and regularization
- What causes the test error? Going beyond bias-variance via ANOVA
- Optimal ridge penalty for real-world high-dimensional data can be zero or negative due to the implicit ridge regularization
- Forward stability and model path selection
- Double data piling: a high-dimensional solution for asymptotically perfect multi-category classification
- Feature importance: a closer look at Shapley values and LOCO
- To bag is to prune
- Analysis and implications of a negative parameter in Tikhonov regularisation
- Why the noise model matters: a performance gap in learned regularization
- Homogenization of SGD in high-dimensions: exact dynamics and generalization properties
- High-dimensional scale invariant discriminant analysis
- The high-dimensional asymptotics of principal component regression
- Improving implicit regularization of SGD with preconditioning for least square problems
- Overparametrized linear regression with noisy and missing data
- Scaling and renormalization in high-dimensional regression
- Augmented balancing weights as linear regression
- Discussion of: ``Augmented balancing weights as linear regression
- Discussion of: ``Augmented balancing weights as linear regression
- Data-driven random projection and screening for high-dimensional generalized linear models
This page was built for publication: The optimal ridge penalty for real-world high-dimensional data can be zero or negative due to the implicit ridge regularization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5148955)
