Private and Interpretable Clinical Prediction with Quantum-Inspired Tensor Train Models

This repository contains the code used to run the experiments in the paper, including training LR models, tensorizing such models, and training DP LR models.

Versions used in the experiments

Python

• python == 3.11.11

Tensor network models

• tensorkrowch == 1.1.5

Differentially private LR models

• diffprivlib == 0.6.6

Machine learning packages

• torch == 2.6.0
• numpy == 2.1.2
• scikit-learn == 1.7.1

Visualization packages

• matplotlib
• seaborn

Note: To render figure texts with $\LaTeX$, ensure that a LaTeX distribution is installed on your system.

Instructions

• Privacy experiments
  See the guide file in the privacy/ folder for instructions on training all model variants and collecting results. All scripts should be executed from the repository’s root directory. Results can be analyzed and visualized in the attacks.ipynb notebook.

• Interpretability experiments
  Conducted in the interpretability.ipynb notebook, located in the interpretability/ folder.