Malware Classification using Deep-Learning

We have used Logistic Regression, Feed Forward NN, CNN, Transfer learning (Inception v3) and LSTM based architectures to demonstrate malware detection (bytecode data preconverted to image data) with an accuracy of 99.6%

Colaborators:

1. Muneeb Ahmed
2. Neda Afreen

This project was carried out at the Department of Computer Engineering @ Jamia Millia Islamia University, New Delhi

---

Dataset for training on ANN, Logistic regression, CNN (Step 1 to 7):

https://drive.google.com/drive/folders/1s7EC4s_-hP9q5vEhs-3vAubspcZbBADK?usp=sharing

Dataset for training on Inception and LSTM (after step 8):

https://drive.google.com/drive/folders/1Nq_xLDDrtsRrNA_WxeDiNvm2o3OO6SQb?usp=sharing

Step 1 to 9 is numbered in the code.

---

* Disclaimer: This project uses Microsoft Big Malware Dataset. Terms and conditions of usage of the dataset are vendored by them.