Efficient NAS with FaDE on Hierarchical Spaces
This work addresses efficiency issues in NAS for researchers and practitioners, offering a method to reduce computational costs in hierarchical search spaces, though it appears incremental as it builds on existing differentiable and evolutionary approaches.
The paper tackles the challenge of neural architecture search (NAS) by introducing FaDE, a method that uses differentiable architecture search to predict relative performance on hierarchical NAS spaces, enabling exploration with linear instead of exponential cost. The experiments show that FaDE ranks correlate with architecture performances and empower an evolutionary search on the complete NAS space.
Neural architecture search (NAS) is a challenging problem. Hierarchical search spaces allow for cheap evaluations of neural network sub modules to serve as surrogate for architecture evaluations. Yet, sometimes the hierarchy is too restrictive or the surrogate fails to generalize. We present FaDE which uses differentiable architecture search to obtain relative performance predictions on finite regions of a hierarchical NAS space. The relative nature of these ranks calls for a memory-less, batch-wise outer search algorithm for which we use an evolutionary algorithm with pseudo-gradient descent. FaDE is especially suited on deep hierarchical, respectively multi-cell search spaces, which it can explore by linear instead of exponential cost and therefore eliminates the need for a proxy search space. Our experiments show that firstly, FaDE-ranks on finite regions of the search space correlate with corresponding architecture performances and secondly, the ranks can empower a pseudo-gradient evolutionary search on the complete neural architecture search space.