PyVertical: A Vertical Federated Learning Framework for Multi-headed SplitNN (2021-04-01T00:00:00.000000Z)

TL;DR

This work introduces PyVertical, a framework supporting vertical federated learning using split neural networks, and presents the training of a simple dual-headed split neural network for a MNIST classification task.

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.

Authors

Daniele Romanini

1 papers

A. Hall

1 papers

Pavlos Papadopoulos

1 papers

TL;DR

Abstract

Authors

References42 items

Federated Learning: Collaborative Machine Learning without Centralized Training Data

Practical Defences Against Model Inversion Attacks for Split Neural Networks

Privacy and Trust Redefined in Federated Machine Learning

Asymmetric Private Set Intersection with Applications to Contact Tracing and Private Vertical Federated Machine Learning

On Deploying Secure Computing: Private Intersection-Sum-with-Cardinality

Private Set Intersection in the Internet Setting From Lightweight Oblivious PRF

SplitNN-driven Vertical Partitioning

FedML: A Research Library and Benchmark for Federated Machine Learning

A Distributed Trust Framework for Privacy-Preserving Machine Learning

SplitFed: When Federated Learning Meets Split Learning

Asymmetrical Vertical Federated Learning

Multi-Participant Multi-Class Vertical Federated Learning

Towards Federated Learning at Scale: System Design

Split learning for health: Distributed deep learning without sharing raw patient data

A generic framework for privacy preserving deep learning

PIR-PSI: Scaling Private Contact Discovery

Distributed learning of deep neural network over multiple agents

Entity Resolution and Federated Learning get a Federated Resolution

Scalable Private Set Intersection Based on OT Extension

Private federated learning on vertically partitioned data via entity resolution and additively homomorphic encryption

Privacy-Preserving Distributed Linear Regression on High-Dimensional Data

Scalable Multi-party Private Set-Intersection

Federated Learning: Strategies for Improving Communication Efficiency

Communication-Efficient Learning of Deep Networks from Decentralized Data

Practical Private Set Intersection Protocols with Linear Complexity

Efficient robust private set intersection

Differential Privacy: A Survey of Results

Privacy-preservation for gradient descent methods

Calibrating Noise to Sensitivity in Private Data Analysis

Building predictors from vertically distributed data

Privacy-preserving distributed mining of association rules on horizontally partitioned data

Privacy preserving regression modelling via distributed computation

Efficient Private Matching and Set Intersection

Privacy-preserving cooperative statistical analysis

Private Matching for Compute

Federated machine learning: Concept and applications

Private Set Intersection: Are Garbled Circuits Better than Custom Protocols?

Privacy-Preserving Analysis of Vertically Partitioned Data Using Secure Matrix Products

Privacy preserving association rule mining in vertically partitioned data

Privacy-Preserving Multivariate Statistical Analysis: Linear Regression and Classification

Source Code

ICLR 2021 - Workshop on Distributed and Private Machine Learning (DPML)

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names