Comparing Train and Test Accuracy in Machine Learning: A Practical Guide

Train vs Test Accuracy Explained (Simple + Practical Guide)

📚 Table of Contents

• Why Model Evaluation Matters
• What is Train Accuracy?
• What is Test Accuracy?
• Overfitting vs Underfitting
• Step-by-Step Example
• Code Example
• CLI Output
• How to Interpret Results
• Key Takeaways
• Related Articles

---

📖 Why Model Evaluation Matters

When building a machine learning model, the goal is NOT just to perform well on known data, but to perform well on new unseen data.

💡 Real goal: Build a model that works in real-world situations, not just on training data.

---

📊 What is Train Accuracy?

Train accuracy tells us how well the model learned from the training data.

If it is high → model has learned patterns well.

BUT…

💡 High train accuracy alone does NOT mean the model is good.

---

🧪 What is Test Accuracy?

Test accuracy checks how well the model performs on new data it has never seen before.

💡 This is the most important metric for real-world performance.

---

⚠️ Overfitting vs Underfitting

Overfitting:

Model memorizes training data but fails on new data.

Underfitting:

Model is too simple and fails on both training and test data.

💡 Think like this:
Overfitting = "memorizing answers"
Underfitting = "not studying enough"

---

📊 Step-by-Step Example

We will use a real dataset: Breast Cancer dataset.

• Features → tumor data
• Target → benign or malignant

---

💻 Code Example

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

# Load data
data = load_breast_cancer()
X = data.data
y = data.target

# Split data
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42
)

# Train model
model = RandomForestClassifier(random_state=42)
model.fit(X_train, y_train)

# Train accuracy
train_pred = model.predict(X_train)
train_acc = accuracy_score(y_train, train_pred)

# Test accuracy
test_pred = model.predict(X_test)
test_acc = accuracy_score(y_test, test_pred)

print("Train Accuracy:", train_acc)
print("Test Accuracy:", test_acc)

---

🖥 CLI Output

Train Accuracy: 1.00
Test Accuracy: 0.96

---

📊 How to Interpret Results

• Train >> Test → Overfitting
• Train ≈ Test (both high) → Good model
• Both low → Underfitting

💡 Ideal case: Train and Test accuracy should be close.

---

🎯 Key Takeaways

✔ Always check test accuracy
✔ Train accuracy alone is misleading
✔ Overfitting is very common
✔ Goal = good performance on unseen data

---

📚 Related Articles

• Train vs Validation vs Test
• Parameter Tuning Guide
• Sentence Categorization
• Softmax vs Probability
• Beyond Accuracy

---

🚀 Final Thought

A good model is not the one that performs best on training data, but the one that performs reliably on new data.