kclip / quantum-CP

Code for the paper "Quantum Conformal Prediction for Reliable Uncertainty Quantification in Quantum Machine Learning"

Quantum Conformal Prediction (QCP)

This repository contains code for "Quantum Conformal Prediction for Reliable Uncertainty Quantification in Quantum Machine Learning" - Sangwoo Park and Osvaldo Simeone.

Dependencies

This program is written in python 3.9.7 and uses PyTorch 1.10.2.

Basic Usage

• All the essential components of QCP can be found in the file 'set_predictors/quantum_conformal_prediction.py'.
• PQC with different angle encodings (fixed, linear, non-linear angle encoding, see Fig. 9) can be found in the file 'quantum_circuit/PQC.py'.
• In order to deploy the above PQC to IBM Quantum NISQ devices, 'quantum_circuit/PQC_with_qiskit.py' might be useful.

Unsupervised Learning (Density Learning for Classical Data)

• Main file is 'main_density_learning.py', while the 'runs/density_learning' folder contains the required running shell scripts.

Supervised Learning (Regression for Classical Data)

• Main file is 'main_regression.py', while the 'runs/regression' folder contains the required running shell scripts.

Quantum Data Classification

• Stand-alone code for quantum data classificaiton can be found in the 'quantum_classification/' folder.