🛡️ Malware Detection Using Hybrid Static + Dynamic Features

---

📌 Overview

This project applies Machine Learning to detect malware by analyzing features extracted from executable files (.exe).
It uses a hybrid feature engineering approach combining:

• 🔹 Static features (header metadata, imports, entropy)
• 🔹 Dynamic behavioral features (runtime indicators if available)

Multiple ML models are trained and evaluated to determine the best-performing approach.

---

📁 Project Structure

File	Description
Malware_comparative_analysis.ipynb	Main notebook containing preprocessing, feature extraction, model training, and visualizations
data_file.csv	Dataset used for the model (must be downloaded separately)
MALWARE PROJECT.docx	Full written report

---

📥 Dataset Requirement

This project requires a dataset to run the notebook.

📌 Download dataset here:
https://www.kaggle.com/datasets/joebeachcapital/windows-malwares After downloading, place it inside the project folder as:

/Malware-Project/data_file.csv

---

🚀 How to Run

Option 1 — Google Colab ✅ (Recommended)

Open Colab:

🔗 https://colab.research.google.com/

Upload the notebook:

Malware_comparative_analysis.ipynb

Upload the dataset when prompted.

---

Option 2 — Local Machine

1️⃣ Clone the repository

git clone https://github.com/rootmass56/Malware-Project.git
cd Malware-Project

2️⃣ Install dependencies

pip install -r requirements.txt

3️⃣ Run Jupyter Notebook

jupyter notebook Malware_comparative_analysis.ipynb

---

🧠 Machine Learning Models Used

Model	Included
Logistic Regression	✔️
Decision Tree	✔️
Random Forest	✔️
Gradient Boosting	✔️
KNN	✔️
LightGBM	✔️ (Best Performance)

Models were evaluated using:

• Accuracy
• Precision
• Recall
• F1-Score

---

🏆 Final Result

📌 LightGBM achieved the highest performance and was the most reliable in detecting malware accurately.

---

📦 Requirements

The project requires the following Python packages:

pandas
numpy
scikit-learn
matplotlib
seaborn
scipy
joblib
lightgbm

A requirements.txt file is included for quick installation.

---

📜 License

This project is intended strictly for research and educational purposes.
Commercial use is not allowed without explicit permission.

---

---

🧾 Citation

If you use this project in research, publications, or academic work, please cite it as:

@software{karpuram2025malware,
  author       = {Saketh Karpuram},
  title        = {Malware Detection Using Hybrid Static and Dynamic Features},
  year         = {2025},
  url          = {https://github.com/rootmass56/Malware-Project},
  note         = {Machine Learning Malware Detection Project},
}

## 👤 Author

**Saketh Karpuram**  

🔗 GitHub: https://github.com/rootmass56  

---

⭐ **If this project helped you, please consider giving it a STAR on GitHub!**