Machine Learning for predictive maintenance of Linux Servers

I’m starting a machine learning project that can predict the next unexpected downtime of linux servers base on collected data of its past performance (CPU, Memory, I/O, etc…).  I’ll be needing any kind of advice and suggestions like architecture, what frameworks to use, what data to collect, etc…

The programming language that I’ll be using is Python3.

This projects is open source so anybody can contribute as well as benefit from it

Hi @athurionThanks for being part of the GitHub Community Forum!

This projects is open source so anybody can contribute as well as benefit from it

If you link to your public repo, other users can more easily find and contribute to the project :smiley:

Cheers!