nc26676027 / private-decision-tree-evaluation

Experimental Code for Private Decision Tree Evaluation

Private Decision Tree Evaluation

This is the artifact for the paper: "Level Up: Private Non-Interactive Decision Tree Evaluation using Levelled Homomorphic Encryption"

Requirements

To build the project, you need the following:

• CMake 3.10 or higher
• C++ compiler with C++17 support
• Microsoft SEAL

Building the project

To build the project

    cd src
    mkdir build
    cd build
    cmake ..
    make
    cmake ../../
    cd src2
    mkdir build
    cd build
    cmake ..
    make

Example to run

Run in the build directory

    ./main -t <path-to-tree> -v <path-to-input> -n <bitlength>

How to run experiments

Assumes the binary is available in src/build.

In experiments/ directory, run

    cd experiments
    ./prepare_data # Prepare data
    python3 run_all_experiments.py # Run all experiments