BayesOpt v0.7.1 released

[Ruben Martinez-Cantin]

This update includes many new developments in Bayesian optimization.
It also improves on the code organization, efficiency and simplifies
the install and configuration process.

The code is such a mature state that it was closed to be v1.0.

More info about BayesOpt and its comparison to related software can be
found in this draft:
http://arxiv.org/abs/1405.7430

Changelog:

BayesOpt 0.7.1

-Added MI criterion
-Simplified Python install
-Fixed bugs in annealed criteria

BayesOpt 0.7

-Refactoring to organize the code.
-MCMC learning for hyperparameter.
-Simpler and uniform interface.
-New configuration parameters (random jump, random seed...)
-New and improved demos.
-Fixed error codes/exceptions in Matlab and Python
-Extended documentation.
-Code optimization.
-Bug fixes.

Best,

Ruben