Synapse Open Synapse
Biology

What makes vertical federated learning work with thirteen samples? A controlled benchmark on coral thermal-stress multi-omics

arXiv q-bio · · 2 min read

arXiv:2601.00050v4 Announce Type: replace Abstract: Vertical federated learning (VFL) lets laboratories that hold different omics layers of the same samples train a joint model without sharing raw data, but it is usually developed for datasets with thousands of samples.

Latest from arXiv q-bio