High-dimensional Bayesian Optimization via Random Projection of Manifold Subspaces (ECML PKDD 2024)
pip install -r requirements.txt
The geometry-aware synthetic experiments contain in the file geometry_aware_synthetic_exp.py
python geometry_aware_synthetic_exp.py --test_func Ackley_Sphere_1 --rep 20 --trial_itr 300 --initial_n 10 --high_dim 500 --effective_dim 10 --proj_dim 15 --update_param 3
The geometry-unaware experiments contain in the file geometry_unaware_synthetic_exp.py
python geometry_unaware_synthetic_exp.py --test_func Ackley_Mix --rep 20 --trial_itr 300 --initial_n 10 --high_dim 500 --effective_dim 15 --proj_dim 15 --update_param 3
The LassoBench experiments contain in the file lasso_exp.py
python lasso_exp.py --test_func Lasso --rep 20 --trial_itr 300 --initial_n 10 --proj_dim 10 --update_param 3
This source code is adopted from: