Code to reproduce the experiments in the paper Semantic variation operators for multidimensional genetic programming

Experiments

Experiments can be run using analysis/ml-analyst/submit_jobs.py. See the command line options (python submit_jobs.py -h) for help.

As example, this command would launch the entire experiment:

python submit_jobs.py --r -ml FeatTuned,FeatSXOTuned,FeatRXOTuned,MLPmod,Linear,XGBoost -n_trials 10 -data ../penn-ml-benchmark/datasets/ -results results/

Notebooks

analysis/ contains these notebooks:

• results_tuning.ipynb produces the tuning results figures.
• results_benchmark.ipynb produces the comparisons to the Where are we now? paper.
• stats.ipynb depends on results_benchmark.ipynb and `results_tuning.ipynb', produces the statistical tests.
• results_benchmark-extended.ipynb contains code to reproduce the extended PMLB results.

Dependencies

• scikit-learn
• xgboost
• datasets come from PMLB
• FEAT