USENIX Security

DataFunSummit

Aug 27, 2023 · Artificial Intelligence

Privacy-Preserving Gradient Boosting Decision Trees via Multi-Party Computation and the Squirrel Framework

This article introduces a privacy-preserving gradient boosting decision tree (GBDT) solution built on multi‑party computation, detailing its background, training steps, the MPC tools used, and the Squirrel framework’s workflow, while discussing performance challenges and experimental results demonstrating scalability to millions of samples.

GBDTMPCSecure Computation

0 likes · 9 min read

Privacy-Preserving Gradient Boosting Decision Trees via Multi-Party Computation and the Squirrel Framework

AntTech

Aug 16, 2023 · Information Security

Ant Group Research Institute Presents Two First-Author Papers at USENIX Security 2023 on Secure MPC for GBDT Training and Efficient 3PC for Binary Circuits

At the 32nd USENIX Security Symposium in Anaheim, Ant Group’s Research Institute sponsored the event and showcased two first‑author papers—one introducing the Squirrel framework for fast, secure two‑party computation of Gradient Boosting Decision Trees, and another proposing an efficient 3‑party protocol for binary circuits in maliciously‑secure DNN inference.

DNN InferenceGradient BoostingMPC

0 likes · 3 min read

Ant Group Research Institute Presents Two First-Author Papers at USENIX Security 2023 on Secure MPC for GBDT Training and Efficient 3PC for Binary Circuits

AntTech

Aug 15, 2023 · Information Security

VILLAIN: Backdoor Attacks Against Vertical Split Learning Presented at USENIX Security 2023

The paper "VILLAIN: Backdoor Attacks Against Vertical Split Learning" introduced at USENIX Security 2023 proposes a novel framework that enables label‑free attackers to infer data labels and inject backdoors into vertically partitioned federated learning models, highlighting new security challenges and defense considerations for collaborative AI systems.

Federated LearningUSENIX Securitybackdoor attack

0 likes · 4 min read

VILLAIN: Backdoor Attacks Against Vertical Split Learning Presented at USENIX Security 2023