الـ Overfitting و Cross-Validation

import pandas as pd
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier

df = pd.read_csv("churn.csv")
df["yearly"] = (df["contract"] == "yearly").astype(int)
features = ["tenure", "bill", "complaints", "data_gb", "yearly"]
X_train, X_test, y_train, y_test = train_test_split(
    df[features], df["churned"], test_size=0.3,
    random_state=42, stratify=df["churned"])

depths = range(1, 21)
train_acc, test_acc = [], []
for d in depths:
    tree = DecisionTreeClassifier(max_depth=d, random_state=42).fit(X_train, y_train)
    train_acc.append(tree.score(X_train, y_train))
    test_acc.append(tree.score(X_test, y_test))

plt.figure(figsize=(7, 3.6))
plt.plot(depths, train_acc, marker="o", label="train", color="#3776AB")
plt.plot(depths, test_acc, marker="o", label="test", color="#E0A800")
plt.xlabel("tree depth")
plt.ylabel("accuracy")
plt.title("Deeper tree: memorizing, not learning")
plt.legend()
plt.grid(alpha=0.3)
plt.show()

import pandas as pd
from sklearn.model_selection import cross_val_score
from sklearn.ensemble import RandomForestClassifier

df = pd.read_csv("churn.csv")
df["yearly"] = (df["contract"] == "yearly").astype(int)
features = ["tenure", "bill", "complaints", "data_gb", "yearly"]

scores = cross_val_score(
    RandomForestClassifier(200, min_samples_leaf=5, random_state=42),
    df[features], df["churned"], cv=5, scoring="roc_auc")

print("Each fold:", scores.round(3))
print(f"Mean AUC: {scores.mean():.3f} (+/- {scores.std():.3f})")

import pandas as pd
from sklearn.model_selection import GridSearchCV
from sklearn.tree import DecisionTreeClassifier

df = pd.read_csv("churn.csv")
df["yearly"] = (df["contract"] == "yearly").astype(int)
features = ["tenure", "bill", "complaints", "data_gb", "yearly"]

grid = GridSearchCV(
    DecisionTreeClassifier(random_state=42),
    param_grid={"max_depth": [2, 3, 4, 5, 6, 8, 12],
                "min_samples_leaf": [1, 5, 20]},
    cv=5, scoring="roc_auc")
grid.fit(df[features], df["churned"])

print("Best settings:", grid.best_params_)
print(f"Best CV AUC: {grid.best_score_:.3f}")