arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 1707.04873cs.LGcs.AI
收藏

基于网络转换的高效体系结构搜索

Efficient Architecture Search by Network Transformation

Shanghai Jiao Tong University(上海交通大学);University College London(伦敦大学学院)

Han Cai, Tianyao Chen, Weinan Zhang, Yong Yu, Jun Wang

摘要:

用于自动设计深度神经网络架构的技术,例如基于强化学习的方法,最近已经显示出有希望的结果。然而,它们的成功是基于巨大的计算资源(例如数百个GPU),这使得它们难以被广泛使用。一个值得注意的局限性是,他们仍然在探索架构空间的过程中从头开始设计和训练每个网络,这是非常低效的。在本文中,我们提出了一个新的框架,以有效的架构搜索探索架构空间的基础上,当前的网络和重用其权重。我们采用了一个强化学习代理作为元控制器,其作用是增加网络深度或层宽度与功能保持转换。因此,以前验证过的网络可以被重用,以进一步探索,从而节省了大量的计算成本。我们应用我们的方法来探索普通卷积神经网络的结构空间(没有跳跃连接,分支等)。图像基准数据集(CIFAR-10,SVHN)上,计算资源有限(5个GPU)。我们的方法可以设计出具有高度竞争力的网络,优于现有的网络使用相同的设计方案。在CIFAR-10上,我们的模型没有跳过连接,达到4.23\%的测试出错率,超过了绝大多数现代架构,接近DenseNet。此外,通过应用我们的方法来探索DenseNet架构空间,我们能够用更少的参数实现更精确的网络。

英文摘要:

Techniques for automatically designing deep neural network architectures such as reinforcement learning based approaches have recently shown promising results. However, their success is based on vast computational resources (e.g. hundreds of GPUs), making them difficult to be widely used. A noticeable limitation is that they still design and train each network from scratch during the exploration of the architecture space, which is highly inefficient. In this paper, we propose a new framework toward efficient architecture search by exploring the architecture space based on the current network and reusing its weights. We employ a reinforcement learning agent as the meta-controller, whose action is to grow the network depth or layer width with function-preserving transformations. As such, the previously validated networks can be reused for further exploration, thus saves a large amount of computational cost. We apply our method to explore the architecture space of the plain convolutional neural networks (no skip-connections, branching etc.) on image benchmark datasets (CIFAR-10, SVHN) with restricted computational resources (5 GPUs). Our method can design highly competitive networks that outperform existing networks using the same design scheme. On CIFAR-10, our model without skip-connections achieves 4.23\% test error rate, exceeding a vast majority of modern architectures and approaching DenseNet. Furthermore, by applying our method to explore the DenseNet architecture space, we are able to achieve more accurate networks with fewer parameters.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑