nahmiasd / DeepSign

Implementation of the DeepSign architecture (https://arxiv.org/pdf/1711.08336.pdf)
3 stars 1 forks source link

Instruction on how to run it #1

Open Alibm73-zz opened 4 years ago

Alibm73-zz commented 4 years ago

Hi @nahmiasd There is no instruction available on how to run this code. I have the JSON file of a malware generated by the cuckoo sandbox (it's called report.json ). what should I do? Could you please make a complete instruction on how to run this? Looking forward to your reply

nahmiasd commented 4 years ago

Hi @Alibm73, The DeepSign class is the keras implementation of the DeepSign architecture (DBN) as specified in the paper. The fit() method is expecting the pre-processed binary TF vectors of the cuckoo reports (as specified in the paper). You can produce these vectors quite easily with scikit-learn's CountVectorizer class.