PyVertical: A Vertical Federated Learning Framework for Multi-headed SplitNN

TL;DR

PyVertical is a framework enabling vertical federated learning with split neural networks, allowing multiple data owners to collaboratively train models without sharing raw data, using Private Set Intersection for entity linking.

Contribution

It introduces PyVertical, a novel framework that supports vertical federated learning with split neural networks and entity linking via Private Set Intersection.

Findings

Successfully trained a dual-headed split neural network on MNIST data

Demonstrated data privacy preservation during collaborative training

Validated framework's effectiveness for vertically partitioned data

Abstract

We introduce PyVertical, a framework supporting vertical federated learning using split neural networks. The proposed framework allows a data scientist to train neural networks on data features vertically partitioned across multiple owners while keeping raw data on an owner's device. To link entities shared across different datasets' partitions, we use Private Set Intersection on IDs associated with data points. To demonstrate the validity of the proposed framework, we present the training of a simple dual-headed split neural network for a MNIST classification task, with data samples vertically distributed across two data owners and a data scientist.