Backprop Evolution
This work addresses the need for improved training efficiency in deep learning, though it appears incremental as it builds on the existing back-propagation paradigm.
The authors tackled the problem of discovering new variations of the back-propagation algorithm, finding several update equations that train faster with short training times than standard back-propagation and perform similarly at convergence.
The back-propagation algorithm is the cornerstone of deep learning. Despite its importance, few variations of the algorithm have been attempted. This work presents an approach to discover new variations of the back-propagation equation. We use a domain specific lan- guage to describe update equations as a list of primitive functions. An evolution-based method is used to discover new propagation rules that maximize the generalization per- formance after a few epochs of training. We find several update equations that can train faster with short training times than standard back-propagation, and perform similar as standard back-propagation at convergence.