Visualización de la curva de aprendizaje con LearningCurveDisplay

• Permite graficar el valor del score vs el tamaño del dataset de entrenamiento.

• Se recomienda su construcción usando from_estimator y no directamente con la función.

Uso directo (no recomendado)

import matplotlib.pyplot as plt
from sklearn.datasets import load_iris
from sklearn.model_selection import LearningCurveDisplay, learning_curve
from sklearn.tree import DecisionTreeClassifier

X, y = load_iris(return_X_y=True)

tree = DecisionTreeClassifier(random_state=0)

train_sizes, train_scores, test_scores = learning_curve(
    tree,
    X,
    y,
)


display = LearningCurveDisplay(
    # -------------------------------------------------------------------------
    # Numbers of training examples that has been used to generate the learning
    # curve.
    train_sizes=train_sizes,
    # -------------------------------------------------------------------------
    # Scores on training sets.
    train_scores=train_scores,
    # -------------------------------------------------------------------------
    # Scores on test set.
    test_scores=test_scores,
    # -------------------------------------------------------------------------
    # The name of the score used in learning_curve. It will be used to decorate
    # the y-axis. If None, the generic name "Score" will be used.
    score_name="Score",
)
display.plot()
plt.show()

Uso de from_estimator()

import matplotlib.pyplot as plt
from sklearn.datasets import load_iris
from sklearn.model_selection import LearningCurveDisplay
from sklearn.tree import DecisionTreeClassifier

X, y = load_iris(return_X_y=True)
tree = DecisionTreeClassifier(random_state=0)
LearningCurveDisplay.from_estimator(tree, X, y)

plt.show()