State-Classifier

A quantum machine learning model for detecting if a Bell diagonal state is separable or not.

Data generation

Bell diagonal states are created with randomised probability coefficients. The separability of the states are checked and saved as a labels for the data points. The features for the ML model are made by taking XX, YY and ZZ Pauli measurements of the states. The features are scaled to the interval [0, 2*Pi].

QML model

The classical features are encoded to the quantum circuit with phase gates and entangling gates. The rest of the quantum circuit acts as a neural network, where weights are updated by a classical optimiser to minimize the loss function. The Pauli Z observable measurements made at the end of the circuit act as the output of the QNN.

To run:

1. Clone the repository.
2. Create and activate a virtual environment.
3. In project root run

   pip install .
   

   to install all the dependencies.
4. Run all of the cells in the model.ipynb notebook.