Code to reproduce the baseline system identification results on several of the nonlinear benchmark datasets available at https://www.nonlinearbenchmark.org/

Accompanying paper explaining identificaiton approaches is available at [preprint link tbc].

All datasets are available to download from the python dataloader.

Results can be run locally by installing requirements and then from run_baselines.py.

For users with HPC access a SLURM compatible submission script is available in run_baselines_slurm.py

Note that some NN models take many hours to train on large datasets.