Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Practical Anonymous Two-Party Gradient Boosting Decision Tree

About

Structured data is well handled by gradient-boosted decision trees (GBDT), which are usually trained on vertically partitioned features across mutually distrustful parties. High speed and interpretability make GBDTs popular in finance and healthcare, where neural networks may fall short. Enabling secure computation for GBDTs poses unique challenges, requiring secure record alignment for comparison. Relying on private set intersection (PSI) is a de facto approach. Mistaking PSI for a safety measure actually exposes which record identifiers (IDs) are shared between the datasets. Although circuit-PSI could help, it is costly for generic uses. New ideas are needed to efficiently train in a "dark forest". Aiming to hide the IDs, we initiate the study of anonymous GBDT training on split data held by two parties. Dual circuit-PSI in our design lets the parties alternate as receiver to run pick-then-sum over local features. Via oblivious programmable pseudorandom functions, we propagate circuit-PSI outputs as shared state across runs. Avoiding universal alignment, we resolve the neglected dilemma that ID hiding incurs a cost that scales with domain size. Next, we halve the cost of ciphertext packing used to convert single-instruction multiple-data homomorphic encryption from (ring) learning with errors in prior secure GBDT (Usenix Security' 23) and related secure machine-learning computations. Comparative experiments show our protocol remains competitive with leaky approaches in efficiency. Enabling ID-hiding aggregation, our techniques can extend to other vertically partitioned analytics.

Huang Chenyu, Zhang Fan, Du Minxin, Chow Sherman SM, Chen Huangxun, Rao Huaming, Huang Danqing, Qian Bo, Chen Peng• 2026

Related benchmarks

TaskDatasetResultRank
ClassificationSKINNONSKIN
F1-Score98.8
7
Binary Classificationa9a
F1 Score65.4
5
Binary Classificationcod-rna
F1 Score88.2
5
Binary ClassificationBreast cancer
F1 Score92
5
Binary ClassificationPhishing
F1 Score95.7
5
Binary Classificationcovtype binary
F1 Score78.2
5
Private GBDT TrainingSynthetic LAN (n=5x10^4, D=4, B=8, m0=8, m1=7)
Training Time (s)39
3
Private GBDT TrainingSynthetic LAN (n=2x10^5, D=4, B=8, m0=8, m1=7)
Training Time (s)96.5
3
Private GBDT TrainingSynthetic LAN (n=1.4x10^5, D=5, B=10, m0=7, m1=16)
Training Time (s)139.1
3
Private GBDT TrainingSynthetic (n=1.4x10^5, D=5, B=10, m0=7, m1=16) - 100Mbps WAN
Training Time (s)668.7
3
Showing 10 of 10 rows

Other info

Follow for update