Lesson 06 � Algorithm

RANDOM FOREST
SABSE STRONG.

Random forest multiple decision trees ka ensemble hai � ek tree galat ho sakta hai, 100 trees nahi. Sabse reliable algorithm hai jo bagging aur feature randomness ka use karta hai.

WHY: Random Forest kyun?

Single decision tree overfit ho sakta hai. Random forest us problem ko solve karta hai � multiple trees banata hai aur unka majority vote leta hai. Sab trees milkar better decision lete hain.

HOW: Random Forest kaise kaam karta hai?

Random Forest ek ensemble learning technique hai jo bagging aur feature randomness ka use karke multiple decision trees banata hai.

Step-by-step process:

1. Bootstrap Sampling: Original dataset se random samples banao (replacement ke saath)
2. Feature Selection: Har tree ko sirf kuch random features do split ke liye
3. Tree Building: Har sample pe apni tree banao
4. Voting: Sab trees ka prediction lo aur majority decide karo

Key Hyperparameters:

• n_estimators: Kitni trees banana hai (default 100)
• max_features: Har split pe kitne features consider karna hai
• max_depth: Tree kitni depth tak ja sakti hai

Code Example

from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
import numpy as np

X = np.random.rand(200, 5)
y = (X[:,0] + X[:,1] > 1).astype(int)

X_train, X_test, y_train, y_test = train_test_split(
 X, y, test_size=0.3, random_state=42
)

model = RandomForestClassifier(n_estimators=100, random_state=42)
model.fit(X_train, y_train)

print(f"Accuracy: {model.score(X_test, y_test):.2%}")
print(f"Feature importances: {model.feature_importances_}")

Try it: code ko run karo

Common mistakes

• Bohot kam trees: Agar n_estimators bahut kam hai (jaise 5) toh ensemble ka fayda nahi milega � 100+ rakhna better hai.
• Feature scaling na karna: Random forest ko feature scaling ki zaroorat nahi hai � yeh tree-based hai.
• Single tree se compare karna: Random forest hamesha single tree se better hota hai � yeh uska main advantage hai.