Noise-Robust Hybrid Quantum Neural Networks Framework
This repository contains the experimental codebase for my Master’s thesis, which investigates the reliability, robustness, and practical limitations of hybrid quantum–classical neural networks in the NISQ era.
This project is structured as a reusable benchmarking framework, not only as a collection of independent demos.
Framework Overview: Key Contribution
The framework uses Qiskit, Qiskit Aer, Qiskit Machine Learning, Cirq, PennyLane, and scikit-learn as underlying computational libraries. The contribution of this project is the reusable evaluation layer built around those libraries.
The framework adds: Standardized dataset handling Synthetic data Iris Wisconsin Diagnostic Breast Cancer (WDBC) Quantum-compatible preprocessing Designed for low-dimensional quantum input Noise-analysis toolbox Depolarizing Bit-flip Phase-flip Amplitude damping Robustness metrics (nonstandard features) accuracy_drop robustness_score degradation_slope training_instability cross_framework_deviation Benchmark pipelines Hybrid vs Classical Noise robustness Cross-framework validation Domain-inspired benchmarks Cybersecurity anomaly detection Medical classification Energy optimization Standardized outputs JSON CSV Accuracy plots Noise curves Heatmaps Framework Structure
Main framework code:
framework/ datasets.py noise_channels.py robustness_metrics.py reporting.py benchmark_runner.py
Main orchestration pipelines:
pipelines/ main_hybrid_vs_classical.py main_noise_robustness.py main_cross_framework_validation.py main_full_benchmark_summary.py main_framework_capabilities_report.py
Framework outputs:
results/framework/ Running the Framework
Run the full framework:
python run_framework.py
Or run pipelines individually:
python -m pipelines.main_hybrid_vs_classical python -m pipelines.main_noise_robustness python -m pipelines.main_cross_framework_validation python -m pipelines.main_full_benchmark_summary python -m pipelines.main_framework_capabilities_report Demonstration Ecosystem
The project includes a 13-demo experimental ecosystem across Qiskit, Cirq, and PennyLane.
Core Demos HQNN Toy Classifier VQE Energy Minimization QAOA MaxCut QSVM Anomaly Detection Noise-Robust HQNN Cross-Framework Noise Benchmark Cross-Platform Parity Consistency HQNN Training Loop (SPSA) Industry-Inspired Demos Medical Risk Classification Energy Grid Optimization Cybersecurity Anomaly Detection HQNN Explainability Cross-Noise Robustness Heatmap Running Individual Demos
Run from the repo root:
python -m demos.core.demo05_hqnn_noise_robust_qiskit
Cybersecurity demo:
python -m demos.industry.demo11_cyber_anomaly_qiskit Environment Setup conda create -n hqnn python=3.11 -y conda activate hqnn pip install -r env/requirements.txt Framework-Specific Features
This framework provides components not available as a unified workflow in default quantum libraries:
accuracy_drop robustness_score degradation_slope training_instability cross_framework_deviation framework-level CSV/JSON reporting cross-framework validation summaries hybrid-versus-classical benchmark summaries Documentation
Includes:
Technical Manuscript Thesis Draft Demo Descriptions Slide Deck Pseudocode Files Status
The repository includes:
Framework layer Pipeline layer Core demos Industry demos Benchmark outputs Contact
GitHub: https://github.com/jeragilo/