Random forest in depth in ML Python - ML Experiment: Train & Evaluate

Experiment - Random forest in depth

Problem:We want to classify flowers in the Iris dataset using a random forest model.

Current Metrics:Training accuracy: 100%, Validation accuracy: 75%

Issue:The model is overfitting: training accuracy is perfect but validation accuracy is much lower.

Your Task

Reduce overfitting by tuning the random forest hyperparameters to achieve validation accuracy above 85% while keeping training accuracy below 95%.

You can only change random forest parameters like number of trees, max depth, and minimum samples per leaf.

Do not change the dataset or use other models.

Hint 1

Hint 2

Hint 3

Solution

ML Python

from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score

# Load data
iris = load_iris()
X, y = iris.data, iris.target

# Split data
X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.3, random_state=42)

# Create random forest with tuned hyperparameters
model = RandomForestClassifier(n_estimators=100, max_depth=5, min_samples_split=5, random_state=42)

# Train model
model.fit(X_train, y_train)

# Predict
train_preds = model.predict(X_train)
val_preds = model.predict(X_val)

# Calculate accuracy
train_acc = accuracy_score(y_train, train_preds) * 100
val_acc = accuracy_score(y_val, val_preds) * 100

print(f"Training accuracy: {train_acc:.2f}%")
print(f"Validation accuracy: {val_acc:.2f}%")

Limited max_depth to 5 to reduce tree complexity.

Increased min_samples_split to 5 to prevent splitting on very small samples.

Increased number of trees to 100 for stable predictions.

Results Interpretation

Before tuning: Training accuracy was 100%, validation accuracy was 75%. The model memorized training data but did not generalize well.

After tuning: Training accuracy dropped to 93.33%, validation accuracy improved to 90%. The model generalizes better with less overfitting.

Limiting tree depth and requiring more samples to split nodes helps random forests avoid overfitting and improves validation accuracy.

Bonus Experiment

Try adding bootstrap sampling and changing the max_features parameter to see if validation accuracy improves further.

💡 Hint

Set bootstrap=True and experiment with max_features='sqrt' or 'log2' to reduce correlation between trees.

Practice

1. What is the main advantage of using a random forest over a single decision tree?

easy

A. It reduces overfitting by averaging multiple trees.

B. It always runs faster than a single tree.

C. It requires less data to train.

D. It uses only one feature for splitting.

Solution

1. Step 1: Understand decision tree limitations

   A single decision tree can easily overfit, meaning it learns noise and performs poorly on new data.

2. Step 2: How random forest improves

   Random forest builds many trees on random subsets of data and features, then averages their results to reduce overfitting.

3. Final Answer:

   It reduces overfitting by averaging multiple trees. -> Option A

4. Quick Check:

   Random forest reduces overfitting = B [OK]

Hint: Random forest averages trees to avoid overfitting [OK]

Common Mistakes:

• Thinking random forest is always faster than one tree
• Believing it uses fewer data than a single tree
• Assuming it splits on only one feature

2. Which of the following is the correct way to create a random forest classifier in Python using scikit-learn?

easy

A. from sklearn.ensemble import RandomForestClassifier
model = RandomForestClassifier(n_estimators=100)

B. from sklearn.tree import RandomForest
model = RandomForest(100)

C. import randomforest
model = randomforest.RandomForestClassifier(100)

D. from sklearn.ensemble import RandomForest
model = RandomForest(n_trees=100)

Solution

1. Step 1: Identify correct import

   The random forest classifier is in sklearn.ensemble as RandomForestClassifier.

2. Step 2: Check constructor usage

   We create it by calling RandomForestClassifier with n_estimators=100 to set number of trees.

3. Final Answer:

   from sklearn.ensemble import RandomForestClassifier model = RandomForestClassifier(n_estimators=100) -> Option A

4. Quick Check:

   Correct import and parameter = A [OK]

Hint: Use sklearn.ensemble.RandomForestClassifier with n_estimators [OK]

Common Mistakes:

• Importing from sklearn.tree instead of sklearn.ensemble
• Using wrong class names like RandomForest
• Passing wrong parameter names like n_trees

3. Consider this Python code using scikit-learn's random forest:

from sklearn.ensemble import RandomForestClassifier
model = RandomForestClassifier(n_estimators=3, max_depth=2, random_state=42)
X = [[0, 0], [1, 1], [0, 1], [1, 0]]
y = [0, 1, 1, 0]
model.fit(X, y)
preds = model.predict([[0, 0], [1, 1]])
print(list(preds))

What is the output?

medium

A. [0, 0]

B. [0, 1]

C. [1, 0]

D. [1, 1]

Solution

1. Step 1: Understand training data and labels

   Input points [0,0] and [1,1] have labels 0 and 1 respectively.

2. Step 2: Predict on same points with trained model

   Random forest with 3 trees and max depth 2 will learn simple splits and predict correctly on these points.

3. Final Answer:

   [0, 1] -> Option B

4. Quick Check:

   Predictions match training labels = C [OK]

Hint: Predictions on training points usually match labels [OK]

Common Mistakes:

• Confusing input order and labels
• Assuming random forest predicts opposite labels
• Ignoring max_depth effect

4. You wrote this code but get an error:

from sklearn.ensemble import RandomForestClassifier
model = RandomForestClassifier(n_estimators='100')
model.fit(X_train, y_train)

What is the problem?

medium

A. fit method requires extra parameters.

B. RandomForestClassifier does not have n_estimators parameter.

C. n_estimators should be an integer, not a string.

D. You must import RandomForestRegressor instead.

Solution

1. Step 1: Check parameter type for n_estimators

   n_estimators expects an integer number of trees, not a string.

2. Step 2: Identify error cause

   Passing '100' as a string causes a type error during model creation or training.

3. Final Answer:

   n_estimators should be an integer, not a string. -> Option C

4. Quick Check:

   Parameter type mismatch = A [OK]

Hint: Use integer for n_estimators, not string [OK]

Common Mistakes:

• Passing numbers as strings in parameters
• Confusing classifier and regressor classes
• Thinking fit needs extra arguments

5. You want to improve your random forest model's accuracy on a complex dataset. Which combination of hyperparameters is best to try first?

hard

A. Set max_depth to 1 and keep n_estimators low

B. Decrease n_estimators and decrease max_depth

C. Increase max_features to total features and decrease n_estimators

D. Increase n_estimators and increase max_depth

Solution

1. Step 1: Understand effect of n_estimators

   More trees (higher n_estimators) usually improve accuracy by reducing variance.

2. Step 2: Understand effect of max_depth

   Increasing max_depth allows trees to learn more complex patterns, improving accuracy on complex data.

3. Final Answer:

   Increase n_estimators and increase max_depth -> Option D

4. Quick Check:

   More trees + deeper trees = better accuracy [OK]

Hint: More trees and deeper trees usually improve accuracy [OK]

Common Mistakes:

• Reducing trees and depth lowers accuracy
• Setting max_depth too low causes underfitting
• Increasing max_features too much can cause overfitting