Offline Training for Memristor-based Neural Networks
Offline Training for Memristor-based Neural Networks
复制标题
基于忆阻器的神经网络的离线训练
DOI：
10.23919/eusipco47968.2020.9287574
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
J. Vicario
中科院分区：
文献类型：
--
作者：
Guillem Boquet;Edwar Macias Toro;A. Morell;Javier Serrano;E. Miranda;J. Vicario
基于硬件神经网络（HNN）的神经形态系统有望成为解决复杂任务的节能计算体系结构。由于所有纳米电子器件都具有共同的可变性，HNN的成功取决于可靠的重量存储或减轻重量变化的技术的发展。在这篇手稿中，我们提出了一种神经网络训练技术，以减轻由于离线学习中权重导入时的电导缺陷而导致的设备到设备变化的影响。为此，我们建议在训练期间将所述变化添加到权重中，以迫使网络学习针对该变化的鲁棒计算。然后，我们实验使用神经网络架构与量化的权重适应忆阻器件所施加的设计约束。最后，我们验证了我们的建议对现实世界的道路交通数据和MNIST图像数据集，分类指标的改进。
Neuromorphic systems based on Hardware Neural Networks (HNN) are expected to be an energy-efficient computing architecture for solving complex tasks. Due to the variability common to all nano-electronic devices, HNN success depends on the development of reliable weight storage or mitigation techniques against weight variation. In this manuscript, we propose a neural network training technique to mitigate the impact of device-to-device variation due to conductance imperfections at weight import in offline-learning. To that aim, we propose to add said variation to the weights during training in order to force the network to learn robust computations against that variation. Then, we experiment using a neural network architecture with quantized weights adapted to the design constrains imposed by memristive devices. Finally, we validate our proposal against real-world road traffic data and the MNIST image data set, showing improvements on the classification metrics.