Kangaroo: A Private and Amortized Inference Framework over WAN for Large-Scale Decision Tree Evaluation
Wei Xu
Network and Distributed System Security (NDSS) Symposium 2026 · Day 2 · Network Security
Wang (presenting on behalf of author Wei Xu, who faced visa issues) introduces Kangaroo, a novel framework for privacy-preserving decision tree inference that achieves **second-level latency** for large-scale tree evaluation over wide area networks. Decision trees and ensemble methods like random forests are widely used in sensitive applications including medical diagnosis, financial risk assessment, and customer behavior prediction. Kangaroo uses **packed homomorphic encryption (PHE)** to amortize computation and communication overhead, achieving **14 to 59x speedup** over state-of-the-art single-round schemes on small datasets while remaining minimally affected by tree structure changes. The framework features constant communication rounds, no offline preprocessing requirements, and support for both client-server and outsourcing scenarios.
AI review
A packed homomorphic encryption framework for private decision tree evaluation. Pure cryptographic engineering with no security relevance beyond the general PETs space. The 14-59x speedup is meaningful for the niche of people building private decision tree inference systems, but this is a systems optimization paper, not security research.