Abstract:Efficient neural architecture search (ENAS) achieves novel efficiency for learning architecture with high-performance via parameter sharing and reinforcement learning (RL). In the phase of architecture search, ENAS employs deep scalable architecture as search space whose training process consumes most of the search cost. Moreover, time-consuming model training is proportional to the depth of deep scalable architecture. Through experiments using ENAS on CIFAR-10, we find that layer reduction of scalable architecture is an effective way to accelerate the search process of ENAS but suffers from a prohibitive performance drop in the phase of architecture estimation. In this article, we propose a broad neural architecture search (BNAS) where we elaborately design broad scalable architecture dubbed broad convolutional neural network (BCNN) to solve the above issue. On the one hand, the proposed broad scalable architecture has fast training speed due to its shallow topology. Moreover, we also adopt RL and parameter sharing used in ENAS as the optimization strategy of BNAS. Hence, the proposed approach can achieve higher search efficiency. On the other hand, the broad scalable architecture extracts multi-scale features and enhancement representations, and feeds them into global average pooling (GAP) layer to yield more reasonable and comprehensive representations. Therefore, the performance of broad scalable architecture can be promised. In particular, we also develop two variants for BNAS that modify the topology of BCNN. In order to verify the effectiveness of BNAS, several experiments are performed and experimental results show that 1) BNAS delivers 0.19 days which is 2.37× less expensive than ENAS who ranks the best in RL-based NAS approaches; 2) compared with small-size (0.5 million parameters) and medium-size (1.1 million parameters) models, the architecture learned by BNAS obtains state-of-the-art performance (3.58% and 3.24% test error) on CIFAR-10; and 3) the learned architecture achieves 25.3% top-1 error on ImageNet just using 3.9 million parameters.

BNAS:An Efficient Neural Architecture Search Approach Using Broad Scalable Architecture

BNAS: Efficient Neural Architecture Search Using Broad Scalable Architecture

Stacked BNAS: Rethinking Broad Convolutional Neural Network for Neural Architecture Search

Binarized Neural Architecture Search

Binarized Neural Architecture Search for Efficient Object Recognition

Evolutionary Neural Architecture Search Combining Multi-Branch ConvNet and Improved Transformer

EAT-NAS: Elastic Architecture Transfer for Accelerating Large-scale Neural Architecture Search

A Fast Evolutionary Knowledge Transfer Search for Multiscale Deep Neural Architecture

AS-NAS: Adaptive Scalable Neural Architecture Search with Reinforced Evolutionary Algorithm for Deep Learning.

Faster Gradient-based NAS Pipeline Combining Broad Scalable Architecture with Confident Learning Rate.

EPNAS: Efficient Progressive Neural Architecture Search

Deeply Supervised Block-Wise Neural Architecture Search

Block Proposal Neural Architecture Search

Simplified Space Based Neural Architecture Search.

SNAS: Stochastic Neural Architecture Search

Efficient Neural Architecture Search via Parameter Sharing

Neural Architecture Search Via Combinatorial Multi-Armed Bandit.

Understanding Architectures Learnt by Cell-based Neural Architecture Search

Toward Evolutionary Multitask Convolutional Neural Architecture Search

Neural Architecture Search with Progressive Evaluation and Sub-Population Preservation

Manas: multi-agent neural architecture search