Spam Detection using Machine Learning

This project implements a spam detection system using classical Machine Learning techniques.
The goal is to classify SMS messages as spam or ham (not spam).

The project compares two baseline models:

• Multinomial Naive Bayes
• Logistic Regression

---

📌 Dataset

The dataset used in this project is the SMS Spam Collection Dataset from UCI:

https://archive.ics.uci.edu/ml/datasets/SMS+Spam+Collection

• ~5,500 labeled SMS messages
• Labels: spam and ham
• Data is automatically downloaded from the source URL during training

---

🧠 Methodology

1. Load dataset from URL
2. Basic data inspection
3. Train/Test split with stratification
4. Feature extraction using TF-IDF
5. Model training:
   • Multinomial Naive Bayes
   • Logistic Regression
6. Model evaluation:
   • Accuracy
   • Precision
   • Recall
   • F1-score
   • Confusion Matrix
7. Threshold tuning for spam probability

---

📊 Results (Example)

Model	Accuracy	Spam Recall
Naive Bayes	~0.98	~0.85
Logistic Regression	~0.98	Higher / Tunable

Logistic Regression provides better flexibility through threshold adjustment, allowing control over precision vs recall trade-off.

---

📁 Project Structure:

spam-classifier-ml/ ├── notebooks/ │ └── spam_classifier.ipynb ├── src/ │ ├── train.py │ ├── evaluate.py │ └── utils.py ├── models/ ├── requirements.txt ├── README.md └── .gitignore