#Failure prediction for machines in a Google cluster

This repository contains the scripts used to perform the data analysis presented in the paper

Towards Data-Driven Autonomics in Data Centers, Alina Sîrbu and Ozalp Babaoglu.

The scripts are released under the GNU General Public Licence v3.0

Files included:

*big_query.sh - commands for BigQuery pre-processing and post-processing of results

*classification.py - python script for training the RF ensemble and evaluating the results