AntiDriftDeepMalware

This repo contains the code of my M.Sc. project:

AntiDriftDeepMalware_MSc_project.pdf

The model trains on dataset created in:

https://github.com/nirosen/Malware-classification-of-the-Berkeley-detection-dataset

How to run:

1. install prequisiotions:

conda env create -f environment.yml && conda activate pytorch17

2. extract the compressed data folder miller60K_80dev_20future.

3. train the model with provided dataset_path (require specification of time_split / rand_split folder)

unzip data.zip

python train.py --dataset_path=data/miller60K_80dev_20future/time_split --train_procedure=all

python train.py --dataset_path=data/miller60K_80dev_20future/rand_split --train_procedure=all

4. anlyze results with jupyter notebooks in analyze_reulsts folder - hypothesis_and_plot and statistic_test.