This project aims to develop a machine learning model that can analyze sensitive user data while ensuring privacy. We implemented techniques such as differential privacy and federated learning to allow for predictive text input and personalized recommendations without compromising user data security.
- Languages: Python
- Libraries: TensorFlow, PyTorch
- Tools: CoreML, Jupyter
data/
: Contains raw and processed data.notebooks/
: Jupyter notebooks for data preprocessing and model training.src/
: Source code for implementing differential privacy, federated learning, and the model.tests/
: Unit tests for the implemented methods.requirements.txt
: Python dependencies.
- Clone the repository:
git clone https://github.com/abhisekjha/privacy-preserving-ml.git cd privacy-preserving-ml
- Install dependencies:
pip install -r requirements.txt
- Preprocess the data:
jupyter notebook notebooks/data_preprocessing.ipynb
- Train the model:
jupyter notebook notebooks/model_training.ipynb
- Enabled secure and private analysis of user data.
- Demonstrated compliance with strict privacy standards.
- Improved user trust and satisfaction with privacy-first features.
This project is licensed under the MIT License.