Extend and improve hyper-parameter tuning

[accosmin]

To implement multi-dimensional hyper-parameter tuning:

• fixed - already implemented, check if the implementation can be simplified
• grid - generalize current log10 to other scales (e.g. linear, log10) and to multiple dimensions
• random - try N random points and report the best configuration

Implement both single and multi threaded versions for each variation.

[accosmin]

Have a look at these pages: http://blog.dato.com/how-to-evaluate-machine-learning-models-part-4-hyperparameter-tuning http://www.johnmyleswhite.com/notebook/2012/07/21/automatic-hyperparameter-tuning-methods/