rmsprop.updater-class: rmsprop updater
In davharris/mistnet: Stochastic Neural Networks

Another updater with adaptive step sizes, like adagrad and adadelta.

https://climin.readthedocs.org/en/latest/rmsprop.html

learning.rate: the learning rate (set to one in the original paper)
squared.grad: a matrix summing the squared gradients over previous updates (decays according to gamma)
decay: how quickly should squared gradients decay?
delta: the delta matrix (see updater)