Resource Constrained Neural Network Architecture Search: Will a Submodularity Assumption Help?

Abstract

The design of neural network architectures is frequently either based on human expertise using trial/error and empirical feedback or tackled via large scale reinforcement learning strategies performed over distinct discrete architecture choices. In the latter case, the optimization is often non-differentiable and also not very amenable to derivative-free optimization methods. Most methods in use today require sizable computational resources. And if we want networks that additionally satisfy resource constraints, the above challenges are exacerbated because the search must now balance accuracy with certain budget constraints on resources. We formulate this problem as the optimization of a set function -- we find that the empirical behavior of this set function often (but not always) satisfies marginal gain and monotonicity principles -- properties central to the idea of submodularity. Based on this observation, we adapt algorithms within discrete optimization to obtain heuristic schemes for neural network architecture search, where we have resource constraints on the architecture. This simple scheme when applied on CIFAR-100 and ImageNet, identifies resource-constrained architectures with quantifiably better performance than current state-of-the-art models designed for mobile devices. Specifically, we find high-performing architectures with fewer parameters and computations by a search method that is much faster.

arXiv commentICCV 2019

References in corpus (12)

• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• Neural Architecture Search with Reinforcement Learning
• AMC: AutoML for Model Compression and Acceleration on Mobile Devices
• Understanding deep learning requires rethinking generalization
• Efficient Neural Architecture Search via Parameter Sharing
• SMASH: One-Shot Model Architecture Search through HyperNetworks
• Efficient Architecture Search by Network Transformation
• Neural Architecture Search with Bayesian Optimisation and Optimal Transport
• Hierarchical Representations for Efficient Architecture Search
• DeepArchitect: Automatically Designing and Training Deep Architectures
• N2N Learning: Network to Network Compression via Policy Gradient Reinforcement Learning
• ANTNets: Mobile Convolutional Neural Networks for Resource Efficient Image Classification

Cited by in corpus (3)

• Bringing AI To Edge: From Deep Learning's Perspective
• TF-NAS: Rethinking Three Search Freedoms of Latency-Constrained Differentiable Neural Architecture Search
• Powering One-shot Topological NAS with Stabilized Share-parameter Proxy