Quantum Neural Network

This repository enables the execution of a Deep Learning algorithm using a Quantum Neural Network.

Update October 2023:

• Changed plot's quality and design.
• Changed way to save results obtained.
• Deleted unused files and requirements.
• Added Confusion Matrix

Usage

Create a shell file by entering the following parameters: dataset name, ephocs, batch size, learning rate and threshold.

#!/bin/bash

python3 main.py -d dataset_name -e epochs_name -b batch_size -r learning_rate -t threshold

After the .sh file is created, run the following command

docker-compose up --build

Files such as plots, confusion matrix, training and validation results, and execution time are saved after execution in a folder called "exp_archive".

Authors

• @Vigimella
• @FrancescoMercaldo
• @Djack1010

If you are using this repository, please cite our work by referring to our publications (BibTex format):

@Article{app122312025,
    AUTHOR = {Mercaldo, Francesco and Ciaramella, Giovanni and Iadarola, Giacomo and Storto, Marco and Martinelli, Fabio and Santone, Antonella},
    TITLE = {Towards Explainable Quantum Machine Learning for Mobile Malware Detection and Classification},
    JOURNAL = {Applied Sciences},
    VOLUME = {12},
    YEAR = {2022},
    NUMBER = {23},
    ARTICLE-NUMBER = {12025},
    URL = {https://www.mdpi.com/2076-3417/12/23/12025},
    ISSN = {2076-3417},
    DOI = {10.3390/app122312025}
    }

Contributing

The authors would like to thank the 'Trust, Security and Privacy' research group within the Institute of Informatics and Telematics (CNR - Pisa, Italy), that support their researches.

In this code we built a Quantum Neural Network (QNN). It is similar to the approach used in Farhi et al

In addition we were inspired by MNIST classification