Code repository for the bachelor thesis: Graph Classification With Simulated Gaussian Boson Sampling

Please note that there are library compatibility issues between the ones needed for GBS simulation and GraKel - used for the classical graph kernels. Therefore, two different environments are used, see: gbs_env.yaml and grakel_env.yaml

An example of the GBS pipeline is given in gbs_example.ipynb

The relevant scripts for each section of the thesis are as follows:

Data - data_processing.py & data_exploration.py

Gaussian Boson Sampling - encoding.py, sampling.py, feature_vector.py & pca.py

Classical Graph Kernels - classical_kernels.py

Machine Learning Classifiers - classical_kernels.py, svm_gbs.py, dummy_classifier.py, MLP.py & GNN.py

Experimental Setup - statistical_testing.py

The pickle files containing the covariance matrices for each dataset were too large to be committed to GitHub.