Add averaging 1D learner

[jbweston]

(original issue on GitLab)

opened by Joseph Weston (@jbweston) at 2017-07-25T15:52:41.859Z

Add a learner that can learn a function that returns a random variable.

This is like the 1D learner problem, but with the additional possibility of re-sampling the function at an already-seen point, x, in order to get a better estimate for the expectation value of f(x).

[jbweston]

originally posted by Anton Akhmerov (@anton-akhmerov) at 2017-07-26T17:24:38.514Z on GitLab

Removing the milestone because this learner is tricky to implement in a robust way.

[jbweston]

originally posted by Bas Nijholt (@basnijholt) at 2018-01-16T13:30:06.039Z on GitLab

@dvarjas maybe we could discuss your ideas here? :)

[jbweston]

originally posted by Jorn Hoofwijk (@Jorn) at 2018-09-25T12:11:36.989Z on GitLab

In what cases would one need this is practice, and do these cases often occur?

[basnijholt]

The work on this feature is done in #143.