https://github.com/jeffbryner/machinelearning

experiments in machine learning
https://github.com/jeffbryner/machinelearning

Last synced: about 2 months ago
JSON representation

experiments in machine learning

• Host: GitHub
• URL: https://github.com/jeffbryner/machinelearning
• Owner: jeffbryner
• License: mpl-2.0
• Created: 2014-03-10T05:20:59.000Z (almost 11 years ago)
• Default Branch: master
• Last Pushed: 2014-03-11T07:04:26.000Z (almost 11 years ago)
• Last Synced: 2024-04-16T03:39:17.346Z (8 months ago)
• Language: Python
• Size: 500 KB
• Stars: 12
• Watchers: 3
• Forks: 4
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
machinelearning

===============

Experiments in machine learning for web logs presented at bsides Vancouver 2014. 

The python notebooks can be run via::

  ipython notebook

in the directory where you place the notebooks. 

The mlTail.py program is meant to take incoming logs via stdin and produce a report of 

bad actors after stdin has stopped. 

These tools all use the excellent topic modelling library: gensim available via::

  pip install gensim

-----

Setup

-----

Ideally you should read the presentation and the ipython notebooks. But if you can't wait:

Steps: 

1) Separate your baseline apache logs into good and bad via the goodFromBad.py script: 

./goodFromBad.py 

2) Send a sample log into mltail.py: 

cat sample.log| ./mltail.py -c options.conf

3) Bask in the glorious output of machine learning telling you who is attacking you

4) Buy me a beer