adaptive priors (add tests)

[davharris]

e.g. EM-based updates

[davharris]

Looks like my EM-based updates are very similar to what David MacKay recommended in his 1995 paper Probable Networks and Plausible Predictions (Equation 26).

He does mean square error and I do variance, though.

[davharris]

See also: Discriminative Transfer Learning with Tree-based Priors (Srivastava and Salakhutdinov) for a useful generalization. Looks like they only update the means, not the variances though.

[davharris]

At minimum, should have machinery to update priors (partly implemented), as well as built-in updates for the Gaussian and Laplace priors.

[davharris]

Added; still need to test.