أشجار القرار والغابة العشوائية

from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import cross_val_score, train_test_split
from sklearn.datasets import make_classification
import pandas as pd
import shap

class="tok-comment"># ── Sample data ────────────────────────────────────────────────────────
X_raw, y = make_classification(n_samples=class="tok-num">500, n_features=class="tok-num">10,
                                n_informative=class="tok-num">5, random_state=class="tok-num">42)
feature_names = [fclass="tok-str">"feature_{i}" for i in range(X_raw.shape[class="tok-num">1])]
X_train, X_test, y_train, y_test = train_test_split(
    X_raw, y, test_size=class="tok-num">0.2, random_state=class="tok-num">42)

class="tok-comment"># Train
rf = RandomForestClassifier(
    n_estimators=class="tok-num">500,
    max_features=class="tok-str">'sqrt',        class="tok-comment"># Random feature subsets
    max_depth=None,             class="tok-comment"># Fully grown (pruned by min_samples_leaf)
    min_samples_leaf=class="tok-num">1,
    oob_score=True,             class="tok-comment"># Free validation estimate
    random_state=class="tok-num">42,
    n_jobs=-class="tok-num">1
)
rf.fit(X_train, y_train)
print(fclass="tok-str">"OOB Score: {rf.oob_score_:.4f}")

class="tok-comment"># Feature importance (Mean Decrease in Impurity)
importances = pd.Series(rf.feature_importances_, index=feature_names)
importances.nlargest(class="tok-num">20).sort_values().plot(kind=class="tok-str">'barh')

class="tok-comment"># SHAP for correct feature importance
explainer = shap.TreeExplainer(rf)
shap_values = explainer.shap_values(X_test)