- Implemented `compute_oof_error_envelope` to calculate absolute prediction-error envelopes and safety margins for adaptive surrogate validation. - Added `build_local_oof_error_table` to create a one-row-per-sample OOF residual table for local safety margins. - Updated `run_nested_cv_rbf_for_target` to utilize the new functions and return validation envelopes and local OOF error data. - Enhanced `main` function to save validation envelopes and local OOF errors to CSV files. - Modified shell scripts to adjust parameters for training and optimization. - Introduced unit tests for new optimization logic, including safety margin calculations and local OOF error table construction.
| Name |
Last commit
|
Last update |
|---|---|---|
| .. | ||
| __pycache__ | Loading commit data... | |
| CHANGELOG_LOCAL_OOF.md | Loading commit data... | |
| README_ACCEPTANCE_CHANGES.md | Loading commit data... | |
| adaptive_acceptance.py | Loading commit data... | |
| de_utils.py | Loading commit data... | |
| flexible_gpr.py | Loading commit data... | |
| flexible_mlp.py | Loading commit data... | |
| ml_optimization_de.py | Loading commit data... | |
| ml_surrogate_train.py | Loading commit data... | |
| rbf_model.py | Loading commit data... | |
| rbf_objective_shap_beeswarm.py | Loading commit data... | |
| rbf_optimization_de.py | Loading commit data... | |
| rbf_surrogate_train.py | Loading commit data... | |
| run_ml.sh | Loading commit data... | |
| run_rbf.sh | Loading commit data... | |
| surrogate_validation.py | Loading commit data... | |
| test_new_optimization_logic.py | Loading commit data... |