AI Paper Atlas

Research map / Learning theory and algorithms / Kernels and regression theory

Stability and L2-penalty in Model Averaging

Hengkun Zhu, Guohua Zou

JMLR 2024 · Learning theory and algorithms › Kernels and regression theory · approach: High-dimensional linear regression theory · 1 citations (OpenAlex)

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 changedWhereMeasuredEffect
add L2-penalty on model weights to Mallows averaging (RM vs MM)Mallows model averaging · wage1 Case 1, n=110mean test MSE: 0.169 -> 0.164Improved
add L2-penalty on model weights to jackknife averaging (RJ vs JM)jackknife model averaging · wage1 Case 1, n=110mean test MSE: 0.168 -> 0.164Improved
add L2-penalty to Mallows averagingMallows model averaging · wage1 Case 2, n=110mean test MSE: 0.158 -> 0.152Improved
add L2-penalty to Mallows averagingMallows model averaging · wage1 Case 2, n=420mean test MSE: 0.146 -> 0.143Improved
add L2-penalty to Mallows averagingMallows model averaging · wage1 Case 1, n=420mean test MSE: 0.148 -> 0.147Improved
use RM instead of AIC model selectionAIC selection · wage1 Case 1, n=110mean test MSE: 0.185 -> 0.164Improved
use RM/RJ instead of smoothed AIC at large nsmoothed AIC · wage1 Case 2, n=420best performance rate: 0.370 -> 0.265Hurt
use RM instead of smoothed AIC at large nsmoothed AIC · wage1 Case 1, n=420best performance rate: 0.175 -> 0.215Hurt
use RJ instead of RM for best performance rateRM · wage1 Case 2, n=110best performance rate: 0.400 -> 0.465Improved
increase sample size n from 110 to 420 for RM/RJ over other methodsRM and RJ · simulations, alpha=0.5, large R2normalized MSEHurt
use RM/RJ instead of other methods in nested simulation with R2 <= 0.5other averaging and selection methods · nested simulation, alpha=0.5normalized MSEImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy