Model Persistence

The physlearn.datasets.google.model_persistence._paper_params module stores the (hyper)parameters used in Boosting on the shoulders of giants in quantum device calibration. Moreover, it provides utilities for retrieving these (hyper)parameters.

References

• Alex Wozniakowski, Jayne Thompson, Mile Gu, and Felix C. Binder. “A new formulation of gradient boosting”, Machine Learning: Science and Technology, 2 045022 (2021).

physlearn.datasets.google.model_persistence._paper_params.paper_params(index=0)

Retrieves a list for StackingRegressor.

Parameters

index (int) – Specifies the single-target regression subtask, using the Python indexing convention.

Returns

params

Return type

list

Examples

>>> from physlearn.datasets import paper_params
>>> paper_params(index=0)
[[{'activation': 'tanh',
   'solver': 'lbfgs',
   'hidden_layer_sizes': (3,),
   'alpha': 17.0,
   'max_iter': 4390},
  {'objective': 'mean_absolute_error',
   'boosting_type': 'goss',
   'num_leaves': 32,
   'max_depth': 20,
   'learning_rate': 0.2,
   'reg_alpha': 0.3,
   'reg_lambda': 0.3,
   'max_bin': 512,
   'subsample_for_bin': 200,
   'n_estimators': 1060}],
 {'activation': 'tanh',
  'solver': 'lbfgs',
  'hidden_layer_sizes': (10,),
  'alpha': 15.0,
  'max_iter': 4070}]

physlearn.datasets.google.model_persistence._paper_params.additional_paper_params(index=0)

Retrieves a dict for LGBMRegressor.

Parameters

index (int) – Specifies the single-target regression subtask, using the Python indexing convention.

Returns

params

Return type

dict

Examples

>>> from physlearn.datasets import additional_paper_params
>>> additional_paper_params(index=0)
{'objective': 'mean_absolute_error',
 'boosting_type': 'goss',
 'num_leaves': 20,
 'reg_alpha': 0.01,
 'reg_lambda': 0.01,
 'n_estimators': 60}

physlearn.datasets.google.model_persistence._paper_params.xgb_paper_params(index=0)

Retrieves a dict for XGBRegressor.

Parameters

index (int) – Specifies the single-target regression subtask, using the Python indexing convention.

Returns

params

Return type

dict

Examples

>>> from physlearn.datasets import xgb_paper_params
>>> xgb_paper_params(index=0)
{'objective': 'reg:squarederror',
 'n_estimators': 80,
 'max_depth': 8,
 'booster': 'dart'}

physlearn.datasets.google.model_persistence._paper_params.supplementary_params(index=0)

Retrieves a dict for MLPRegressor.

Parameters

index (int) – Specifies the single-target regression subtask, using the Python indexing convention.

Returns

params

Return type

dict

Examples

>>> from physlearn.datasets import supplementary_params
>>> supplementary_params(index=0)
{'activation': 'relu',
 'solver': 'lbfgs',
 'hidden_layer_sizes': (10,),
 'alpha': 15.0,
 'max_iter': 4600}