Hybrid ML-driven framework to predict and mitigate noise in quantum circuits using Random Forest and MLP regressors. Generates, simulates, and analyzes quantum circuits under noisy conditions.
QNoiseNet is a Python-based project that combines quantum computing simulations and machine learning techniques to predict the impact of noise on quantum circuits. It:
- Generates random quantum circuits with variable qubits and gate configurations.
- Simulates quantum circuits using Qiskit Aer, both under ideal and noisy conditions.
- Creates a dataset of circuit features and output probability distributions.
- Trains Random Forest and MLP regressors to predict noise-free probability distributions from noisy simulations.
- Evaluates model performance using metrics like Mean Squared Error (MSE) and Fidelity.
- Visualizes probability distributions and performance comparisons.
- Random quantum circuit generator (2–5 qubits)
- Noisy circuit simulation with customizable noise models
- Dataset generation and storage (
training_data.csv) - ML training (Random Forest & MLP) and prediction
- Fidelity and MSE-based performance evaluation
- Probability distribution visualization
git clone https://github.com/yourusername/QNoiseNet.git
cd QNoiseNet
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
pip install -r requirements.txt