Network Cross-Validation and Model Selection via Subsampling

Sayan Chakrabarty · 2025

Read Network Cross-Validation and Model Selection via Subsampling by Sayan Chakrabarty (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-06903.txt.

Introduction

Sayan Chakrabarty, Srijan Sengupta, and Yuguo Chen propose a cross validation method for network data in a 2025 paper. Checking a model on held out data is standard practice, but networks resist the usual split since every node pair links to others. Bad validation leads to poor model choice in fields from biology to social science.

Their method cuts the network into overlapping parts that share a common core. Models train on the parts and are tested on the pairs that fall between them, which keeps the split honest. The smaller training pieces also cut computing cost on large networks. Theory backs the method for model choice and tuning tasks, and tests show it often beats past network validation on both accuracy and speed.

Catalog record

CREATORSayan Chakrabarty
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE180 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-06903

Browse the full library · About the Tower of Babel