[R] Has there been any work on alternating minimization methods for training neural nets?

Open in new window