Parameters of PyCaret modeling functions

Created at : 2024-10-31 14:04
Auther: Soo.Y

---

📝메모

compare_models

실행

compare_models은 다양한 모델을 훈련한 후에 성능을 비교해서 best model를 선정해 줍니다.

s.compare_models(
    include=None,
    exclude=None,
    fold=None,
    round=4,
    cross_validation=True,
    sort="Accuracy",
    n_select=1,
    budget_time=None,
    turbo=True,
    errors="ignore",
    fit_kwargs=None,
    groups=None,
    experiment_custom_tags=None,
    probability_threshold=None,
    engine=None,
    verbose=True,
    parallel=None,
)

파라미터

• include
  • defalut = None
  • list of str
  • 사용하고자 하는 모델을 입력
  • 모델 라이브러리 사용 가능
• exclude
  • defalut = None
  • list of str
  • 제외하고자 하는 모델을 입력
  • 모델 라이브러리 사용 가능
• fold
  • defalut = None
  • int or scikit-learn CV generator
  • cross-validation의 수를 입력
• round : 자릿수, 기본 4
• sort : 정렬, deault = “Accuracy”
• n_select
  • defalut = 1
  • top model의 수

create_model

실행

s.create_model(
    estimator,
    fold=None,
    round=4,
    cross_validation=True,
    fit_kwargs=None,
    groups=None,
    experiment_custom_tags=None,
    probability_threshold=None,
    engine=None,
    verbose=True,
    return_train_score=False,
    **kwargs,
)

파라미터

• estimator(str or scikit-learn compatible object)
  • lr : Logistic Regression
  • knn : K Neighbors Classifier
  • nb : Naive Bayes
  • dt : Decision Tree Classifier
  • svm : SVM - Linear Kernel
  • rbfsvm : SVM - Radial Kernel
  • gpc : Gaussian Process Classifier
  • mlp : MLP Classifier
  • ridge : Ridge Classifier
  • rf : Random Forest Classifier
  • qda : Quadratic Discriminant Analysis
  • ada : Ada Boost Classifier
  • gbc : Gradient Boosting Classifier
  • lda : Linear Discriminant Analysis
  • et : Extra Trees Classifier
  • xgboost : Extreme Gradient Boosting
  • lightgbm : Light Gradient Boosting Machine
  • catboost : CatBoost Classifier
• fold
  • 교차 검증 수

tune_model

실행

s.tune_model(
    estimator,
    fold=None,
    round=4,
    n_iter=10,
    custom_grid=None,
    optimize="Accuracy",
    custom_scorer=None,
    search_library="scikit-learn",
    search_algorithm=None,
    early_stopping=False,
    early_stopping_max_iters=10,
    choose_better=True,
    fit_kwargs=None,
    groups=None,
    return_tuner=False,
    verbose=True,
    tuner_verbose=True,
    return_train_score=False,
    **kwargs,
)

파라미터

• estimaotr : 훈련된 모델 오브젝트
• fold : 교차 검증 수
• n_iter : grid search의 반복 횟수
• custom_grid : custom search space for hyperparameters
• optimize : “Accuracy”
• search_library
  • str, default = “scikit-learn”
  • ‘scikit-learn’ - default, requires no further installation              https://github.com/scikit-learn/scikit-learn            ‘scikit-optimize’ - pip install scikit-optimize              https://scikit-optimize.github.io/stable/           ‘tune-sklearn’ - pip install tune-sklearn ray[tune]              https://github.com/ray-project/tune-sklearn            ‘optuna’ - pip install optuna             https://optuna.org/
• search_algorithm
  • str, defalut = None
  • 종류
    • scikit-learn : ‘random’, ‘grid’
    • scikit-optimize : ‘bayesian’
    • tune-sklearn : ‘random’, ‘grid’, ‘bayesian’, ‘hyperopt’, ‘optuna’, ‘bohb’
    • optuna : ‘random’, ‘tpe’
• early_stopping
  • defalut = False
  • (주의) serach_library에서 scikit-learn을 사용하면 early_stopping 사용 불가
  • ‘asha’ : Asnchronous Succesive Halving Algorithm
  • ‘hyperband’ : Hyperband
  • ‘median’ : Median Stopping Rule
• early_stopping_max_iters : int, defalut = 10

ensemble_model

실행

s.ensemble_model(
        estimator=, # Trained model object
        method="Bagging",
        fold=None,
        n_estimators=10,
        round=4,
        choose_better=False,
        optimize="Accuracy",
        fit_kwargs=None,
        groups=None,
        probability_threshold=None,
        verbose=True,
        return_train_score=False,
    )

파라미터

• estimator : scikit-learn or Trained model object
• method
  • ‘Bagging’ or “Boosting”
• fold : 교차 검증 수
• n_estimators
  • default = 10
  • ensemble의 기본 estimator의 수
• optimize : ‘Accuracy’

blend_models

실행

s.blend_models(
        estimator_list, # list,
        fold=None,
        round=4,
        choose_better=False,
        optimize="Accuracy",
        method="auto",
        weights=None,
        fit_kwargs=None,
        groups=None,
        probability_threshold=None,
        verbose=True,
        return_train_score=False,
    )

파라미터

• estimator_list : List of trained model objects
• fold : 교차 검증 수
• optimize : ‘Accuracy’

stack_models

실행

s.stack_models(
    estimator_list: list,
    meta_model=None,
    meta_model_fold=5,
    fold=None,
    round=4,
    method="auto",
    restack=False,
    choose_better=False,
    optimize="Accuracy",
    fit_kwargs=None,
    groups=None,
    probability_threshold=None,
    verbose=True,
    return_train_score=False,
)

파라미터

• estimaotr_list
• meta_model
• fold : 교차 검증 수
• optimize : ‘Accuracy’

---

📜출처(참고 문헌)

---

🔗연결 문서