A Theory of Feature Learning in Kernel Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Yunlu, Li, Yang, Liu, Keli, Ruan, Feng
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908834734276608
author Chen, Yunlu
Li, Yang
Liu, Keli
Ruan, Feng
author_facet Chen, Yunlu
Li, Yang
Liu, Keli
Ruan, Feng
contents We study feature learning in a compositional variant of kernel ridge regression in which the predictor is applied to a learnable linear transformation of the input. When the response depends on the input only through a low-dimensional predictive subspace, we show that all global minimizers of the population objective for the linear transformation annihilate directions orthogonal to this subspace, and in certain regimes, exactly identify the subspace. Moreover, we show that global minimizers of the finite-sample objective inherit the exact same low-dimensional structure with high probability, even without any explicit penalization on the linear transformation.
format Preprint
id arxiv_https___arxiv_org_abs_2310_11736
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A Theory of Feature Learning in Kernel Models
Chen, Yunlu
Li, Yang
Liu, Keli
Ruan, Feng
Statistics Theory
Optimization and Control
Machine Learning
We study feature learning in a compositional variant of kernel ridge regression in which the predictor is applied to a learnable linear transformation of the input. When the response depends on the input only through a low-dimensional predictive subspace, we show that all global minimizers of the population objective for the linear transformation annihilate directions orthogonal to this subspace, and in certain regimes, exactly identify the subspace. Moreover, we show that global minimizers of the finite-sample objective inherit the exact same low-dimensional structure with high probability, even without any explicit penalization on the linear transformation.
title A Theory of Feature Learning in Kernel Models
topic Statistics Theory
Optimization and Control
Machine Learning
url https://arxiv.org/abs/2310.11736