New GP model

[ric70x7]

We should integrate EP_GP and GP_regression models into a single one. That way it will be easier to keep them both up to date.

Since the log marginal likelihood for an EP model can be written as the log likelihood of a regression model for a new variable Y* = v_tilde/tau_tilde, with a covariance matrix K* = K + diag(1./tau_tilde) plus a normalization term, we can use most of the GP_regression code and just add other functions to call the EP algorithm.

Then we can also implement sparse _GP_regression and sparse_EP_GP into the same model.

For consistency between the GP_regression and the sparse_GP_regression, and also to make more clear the differences with EP, in the non-sparse regression, beta should be explicit rather than part of the kernel.

I'll open a branch called newGP for this.

[jameshensman]

The EPEM algorithm does not belong in model, TODO: move it to the GP class or similar

[jameshensman]

Okay, after discussion, the EPEM algorithm should clearly live in the inference and optimisation directory

[jameshensman]

Done. EP and non-EP models now work consistently. There's also room in the framework for other likelihood approximations (Laplace and VB spring to mind).