This code regenerates the results obtained in the paper Hyperparameter Importance of Quantum Neural Networks Across Small Datasets. This work has also been extended to include an extra set of experiments assessing the usefulness of using priors of good hyperparameter values of Quantum Neural Networks (QNNs) for hyperparameter optimization. We provide the performance data obtained when running QNNs in csv format. The notebooks can be run to obtain the resulting plots of the verification experiment (random search with surrogate models) and the extended set of experiments for hyperparameter optimization.
Run notebook verification-experiment-fanova-surrogate.ipynb.
Run notebook prior-good-params-experiment-fanova-surrogate.ipynb.