OpenFHE - SVM

OpenFHE-Python-based Examples of Encrypted Support Vector Machine Inference based on this C++ example.

Preparing the environment

Before running any code from this repo, make sure you have a working Python 3.6+ installation. Then, install the dependencies:

1 - PyBind11:

pip install "pybind11[global]" # or alternatively, if you use conda:
conda install -c conda-forge pybind11

2 - OpenFHE-development:

git clone git@github.com:openfheorg/openfhe-development.git
cd openfhe-development
mkdir build
cd build
cmake ..
make -j N  # number of processors - 4 in my case
sudo make install && cd ../.. # to get back to original path

3 - OpenFHE-Python:

git clone git@github.com:openfheorg/openfhe-python.git
cd openfhe-python
mkdir build
cd build
cmake .. # or alternatively for conda: cmake .. -DPYTHON_EXECUTABLE_PATH=$CONDA_PREFIX/bin/python
make -j N  # number of processors - 4 in my case
sudo make install && cd ../.. # to get back to original path

4 - Model Requirements:

git clone git@github.com:reneroliveira/openfhe-svm.git
cd openfhe-svm
pip install -r requirements.txt

Running

First, install openfhe-python following the instructions of library's repository.

Then, you can run python model_training.py to train and save the model weights, and then call the encrypted model files for inference. All necessary data is inside data/ directory, but if you want to reproduce the data generation run the get_data.py file.

The main files are encrypted_svm_linear.py and encrypted_svm_poly.py, which are the encrypted versions of inference of the linear and polynomial kernel SVM models, respectively.