Analyzing the Short-Term Dependency in Ultra-High Magnetic Response Systems - Modeling Sequential Data with Non-Recurrent Neural Networks

Analyzing the Short-Term Dependency in Ultra-High Magnetic Response Systems - Modeling Sequential Data with Non-Recurrent Neural Networks
复制标题

分析超高磁响应系统的短期依赖性 - 使用非循环神经网络建模序列数据

DOI:
10.1016/j.procs.2021.05.044
复制
发表时间:
2021
期刊:
Procedia Computer Science
影响因子:
--
通讯作者:
Li, Lichun
Li, Lichun
中科院分区:
--
文献类型:
--
作者:
Sun, Jieming;Li, Lichun

文献摘要

参考文献

相似文献

循环神经网络 (RNN) 是一种流行的顺序数据建模选择。然而,经验表明 RNN 的调整和定制通常很困难且耗时。它促使从业者用非循环神经网络取代 RNN,后者在某些情况下表现出相当的性能。使用非循环神经网络对序列数据进行建模的成功表明了序列数据之间的短期依赖性。在本文中,我们基于动态系统的可观测性准则,系统地分析了具有部分系统知识的超高磁响应系统(UHMR)的短期依赖性。此外,我们表明 UHMR 系统中的顺序数据仅具有两步依赖性。该结果表明实验中任何连续的三个步骤都可以形成训练前馈神经网络(FFNN)的数据。因此,可以通过少量的实验收集足够的数据。基于分析,我们训练前馈神经网络,根据四个实验的连续数据对 UHMR 系统进行建模。通过适当的数据预处理,FFNN 模型可以以有界平均绝对误差预测系统性能。
Recurrent neural network (RNN) is a popular modeling choice for sequential data. However, empirical experience shows RNNs are often difficult and time-consuming to tune and customize. It drives practitioners to replace RNNs with non-recurrent neural networks which presented comparable performances in some cases. The success of using non-recurrent neural networks to model sequential data indicates the short-term dependency among sequential data. In this paper, we systematically analyze the short-term dependency in ultra-high magnetic response systems (UHMR) with partial system knowledge based on the observability criterion of the dynamic systems. Moreover, we show that the sequential data in the UHMR system only have 2-step dependency. This result indicates that any consecutive three steps in an experiment form a datum to train a feed-forward neural network (FFNN). Therefore, sufficient data can be collected within a small number of experiments. Based on the analysis, we train a feed-forward neural network to model the UHMR system based on the sequential data from four experiments. Through proper data pre-processing, the FFNN model can predict the system performance with bounded mean absolute error.
生物多样性、生命之树和科学传播
DOI: 10.1007/978-3-319-93145-6_3
发表时间: 2018
影响因子: 6.8
作者:
J. Rosindell;Yan Wong
通讯作者: Yan Wong
离散时间非线性系统的类 Luenberger 观测器
DOI: 10.1109/cdc.1998.762057
发表时间: 1997
期刊: 1997 European Control Conference (ECC)
影响因子: --
作者:
S. Benamor;H. Hararaouri;F. Couenne
通讯作者: F. Couenne
离散时间非线性系统一致可观性的进一步结果
DOI: 10.1109/tac.2010.2041983
发表时间: 2010
影响因子: 6.8
作者:
S. Hanba
通讯作者: S. Hanba
用于非线性系统建模和控制的动态神经网络
DOI: 10.1080/10798587.2000.10642843
发表时间: 2003
期刊: Intell. Autom. Soft Comput.
影响因子: --
作者:
F. Pourboghrat;H. Pongpairoj;Ziqian Liu;F. Farid;F. Pourboghrat;B. Aazhang
通讯作者: B. Aazhang
离散时间非线性系统的 Luenberger 观测器
DOI: 10.1109/cdc40024.2019.9029220
发表时间: 2019
期刊: 2019 IEEE 58th Conference on Decision and Control (CDC)
影响因子: --
作者:
L. Brivadis;V. Andrieu;U. Serres
通讯作者: U. Serres