A Novel Framework for Neural Architecture Search in the Hill Climbing Domain
This addresses the need for efficient neural architecture search to reduce manual expertise and computational costs, though it appears incremental as it builds on existing hill-climbing methods.
The paper tackles the problem of automating neural architecture design by proposing a hill-climbing framework with morphism operators and a gradient update based on layer aging, resulting in a 4.96% error rate on CIFAR-10 in 19.4 hours on a single GPU.
Neural networks have now long been used for solving complex problems of image domain, yet designing the same needs manual expertise. Furthermore, techniques for automatically generating a suitable deep learning architecture for a given dataset have frequently made use of reinforcement learning and evolutionary methods which take extensive computational resources and time. We propose a new framework for neural architecture search based on a hill-climbing procedure using morphism operators that makes use of a novel gradient update scheme. The update is based on the aging of neural network layers and results in the reduction in the overall training time. This technique can search in a broader search space which subsequently yields competitive results. We achieve a 4.96% error rate on the CIFAR-10 dataset in 19.4 hours of a single GPU training.