Unsupervised ML Intrusion detection system - Printable Version +- Python Forum (https://python-forum.io) +-- Forum: Python Coding (https://python-forum.io/forum-7.html) +--- Forum: Data Science (https://python-forum.io/forum-44.html) +--- Thread: Unsupervised ML Intrusion detection system (/thread-25280.html) |
Unsupervised ML Intrusion detection system - CammyS32 - Mar-25-2020 Hey guys, I am new to coding. I study Digital Forensics, security and ethical hacking at uni in which I want to pursue Pen testing. For my 4th year dissertation I am trying to make an Intrusion detection system that utilises unsupervised machine learning, k-means clustering. I have an industry standard dataset made up of begnin and network attack pcap's, fed into jupyter notebook, which I want to be clustered into Begnin traffic or attacks. Github code: https://github.com/CammyS32/Project (PractiseHonsProject) Datasheet: https://www.dropbox.com/s/kn3w77zgnjhxty9/Data_file_4.csv.csv?dl=0 So far I have some code working however I don't understand the output of my code. The scatter plot doesn't look like other k-means scatter plots and I can't visualise how I am going to evaluate the output. I am hoping you can shed some light on the output of my code and some next steps. Any tips on how to improve my code are welcomed too Tia, Cammy |