CatBoost¶
CatBoost is a state-of-the-art open-source gradient boosting on decision trees library. eli5 supports eli5.explain_weights()
for catboost.CatBoost
, catboost.CatBoostClassifier
and catboost.CatBoostRegressor
.
eli5.explain_weights()
uses feature importances. Additional
arguments for CatBoostClassifier and CatBoostRegressor:
importance_type
is a way to get feature importance. Possible values are:- ‘PredictionValuesChange’ - The individual importance values for each of the input features.(default)
- ‘LossFunctionChange’ - The individual importance values for each of the input features for ranking metrics (requires training data to be passed or a similar dataset with Pool)
pool
thecatboost.Pool
datatype . To be passed ifexplain_weights_catboost
has importance_type set to ‘LossFunctionChange’. The catboostfeature_importances
uses thePool
datatype to calculate the parameter for the specificimportance_type
.
Note
Top-level eli5.explain_weights()
calls are dispatched
to eli5.catboost.explain_weights_catboost()
for
catboost.CatBoost
, catboost.CatBoostClassifer
and catboost.CatBoostRegressor
.