SplitNN-driven Vertical Partitioning

By Iker Ceballos, Vivek Sharma, E. Múgica, Abhishek Singh, Alberto Roman, Praneeth Vepakomma, Ramesh Raskar. 2020

91 citations recorded by the source corpus.

In this work, we introduce SplitNN-driven Vertical Partitioning, a configuration of a distributed deep learning method called SplitNN to facilitate learning from vertically distributed features. SplitNN does not share raw data or model details with collaborating institutions. The proposed configuration allows training among institutions holding diverse sources of data without the need of complex encryption algorithms or secure computation protocols. We evaluate several configurations to merge the outputs of the split models, and compare performance and resource efficiency. The method is flexible and allows many different configurations to tackle the specific challenges posed by vertically split datasets.

Subfields: Federated & Distributed Learning, Privacy & Security

Citations in the corpus

• Vertical Federated Learning: Concepts, Advances, and Challenges (2022, unknown): “If a trainable global module is in place, this VFL scenario is coincident with the vertical splitNN [32], where the whole model is split into different parties, thus we term it splitVFL (see Fig.”