AI-GCS: Execute evaluation and tuning of algorithms

[pombredanne]

This is to design and execute a first set of evaluation tests of the
algorithm at a decent enough scale to tune the performance of the
algorithms. And to deploy an initial evaluation infrastructure. The tasks are:

• Deploy servers to run evaluations
• Create and public report from evaluation runs
• Create scripts to execute evaluation
• Update open source library