An Agnostic View on the Cost of Overfitting in (Kernel) Ridge Regression
Authors: Lijia Zhou, James B Simon, Gal Vardi, Nathan Srebro
ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | We study the cost of overfitting in noisy kernel ridge regression (KRR)... We analyze the cost of overfitting under a Gaussian universality ansatz using recently derived (non-rigorous) risk estimates in terms of the task eigenstructure. Our analysis provides a more refined characterization of benign, tempered and catastrophic overfitting (cf. Mallinar et al., 2022).For the rest of this paper, we will simply treat it as an equivalence, formally proving facts about the omniscient risk estimate R( ˆfδ). Thus, our results follow by analyzing the expression from (6). |
| Researcher Affiliation | Collaboration | Lijia Zhou University of Chicago zlj@uchicago.edu James B. Simon UC Berkeley and Generally Intelligent james.simon@berkeley.edu Gal Vardi TTI-Chicago and Hebrew University galvardi@ttic.edu Nathan Srebro TTI-Chicago nati@ttic.edu |
| Pseudocode | No | The paper does not contain any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any concrete access to source code for the methodology described. |
| Open Datasets | No | The paper is theoretical and does not use or describe any datasets for experimental purposes, thus no public access information for datasets is provided. |
| Dataset Splits | No | The paper is theoretical and does not report experimental results requiring training/validation/test splits. |
| Hardware Specification | No | The paper is theoretical and does not report computational experiments that would require specific hardware specifications. |
| Software Dependencies | No | The paper is theoretical and does not report computational experiments that would require specific software dependencies with version numbers. |
| Experiment Setup | No | The paper is theoretical and does not detail experimental setups such as hyperparameters or training configurations. |