Self-Distillation is Optimal Among Spectral Shrinkage Estimators in Spiked Covariance Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lecoiu, Radu, Mukherjee, Debarghya, Sur, Pragya
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910229698969600
author Lecoiu, Radu
Mukherjee, Debarghya
Sur, Pragya
author_facet Lecoiu, Radu
Mukherjee, Debarghya
Sur, Pragya
contents Self-distillation has emerged as a promising technique for improving model performance in modern machine learning systems. We develop the statistical foundations of self-distillation in spiked covariance models, by introducing and analyzing a broad class of estimators, namely spectral shrinkage estimators. We establish that for spiked covariance matrices with $s$ spikes, $s$-step self-distillation achieves optimal performance among spectral shrinkage estimators, outperforming well-known estimators in statistics and machine learning. Moreover, we show that $s$ steps are necessary for optimality: any $(s-k)$-step distilled estimator is strictly suboptimal for $1 \leq k \leq s$. For the special subclass of isotropic covariances, we show that optimally tuned Ridge regression performs best among spectral shrinkage estimators. We also study a federated approach where multiple data centers share spectral shrinkage estimators and a common server seeks to aggregate them to achieve optimal performance. In this case, we find that the best local rule again takes the form of self-distillation, though it differs from the optimal rule when data are hosted centrally on a single server. Together, our results elucidate why self-distillation improves predictive performance and provide a broader statistical framework connecting it with classical shrinkage-based methods.
format Preprint
id arxiv_https___arxiv_org_abs_2605_17778
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Self-Distillation is Optimal Among Spectral Shrinkage Estimators in Spiked Covariance Models
Lecoiu, Radu
Mukherjee, Debarghya
Sur, Pragya
Statistics Theory
Machine Learning
Methodology
Self-distillation has emerged as a promising technique for improving model performance in modern machine learning systems. We develop the statistical foundations of self-distillation in spiked covariance models, by introducing and analyzing a broad class of estimators, namely spectral shrinkage estimators. We establish that for spiked covariance matrices with $s$ spikes, $s$-step self-distillation achieves optimal performance among spectral shrinkage estimators, outperforming well-known estimators in statistics and machine learning. Moreover, we show that $s$ steps are necessary for optimality: any $(s-k)$-step distilled estimator is strictly suboptimal for $1 \leq k \leq s$. For the special subclass of isotropic covariances, we show that optimally tuned Ridge regression performs best among spectral shrinkage estimators. We also study a federated approach where multiple data centers share spectral shrinkage estimators and a common server seeks to aggregate them to achieve optimal performance. In this case, we find that the best local rule again takes the form of self-distillation, though it differs from the optimal rule when data are hosted centrally on a single server. Together, our results elucidate why self-distillation improves predictive performance and provide a broader statistical framework connecting it with classical shrinkage-based methods.
title Self-Distillation is Optimal Among Spectral Shrinkage Estimators in Spiked Covariance Models
topic Statistics Theory
Machine Learning
Methodology
url https://arxiv.org/abs/2605.17778