时间历史术语对水库动态和回声状态网络中预测准确性的影响
Yudai Ebato1, Sou Nobukawa2,3,4,5, Yusuke Sakemi5,6
1Graduate School of Information and Computer Science, Chiba Institute of Technology, 2-17-1 Tsudanuma, Narashino, Chiba, 275-0016, Japan. yudaiebato@gmail.com.
具有时间历史术语的回声状态网络 (ESN) 显示了改进的时间序列预测. 这种增强与延迟能力的增加,保持水库多样性和稳定性有关,以提高性能.
科学领域:
- 机器学习 机器学习
- 计算神经科学是一种神经科学.
- 动态系统 动态系统
背景情况:
- 由于高效的训练,回声状态网络 (ESN) 对于时间序列处理是有效的.
- 具有储库的循环神经网络是ESN的核心,处理输入信号.
- 储存神经元中的时间历史术语可以改善ESN预测,但缺乏定量解释.
研究的目的:
- 用时间历史术语量化解释ESN的性能提升.
- 研究延迟能力在ESN性能中的作用.
- 为了比较ESN的储动态,有和没有时间历史条款.
主要方法:
- 使用漏洞集成器ESN (LI-ESN) 和混乱回声状态网络 (ChESN) 的比较实验.
- 分析水库动态特征,包括多样性,稳定性和延迟能力.
- 评估时间序列预测性能.
主要成果:
- 具有时间历史条款 (LI-ESN,ChESN) 的 ESN 与标准 ESN 相比表现优越.
- LI-ESN和ChESN的水库动态保持了多样性和稳定性.
- 在LI-ESN和ChESN中观察到更高的延迟能力,与性能改善相关.
结论:
- 具有时间历史术语的ESN的性能提升归因于延迟能力的增加.
- 以多样性,稳定性和高延迟能力为特征的水库动态是ESN卓越性能的关键.
- 本研究为评估和开发ESN架构提供了一个动态系统的视角.
更多相关视频
10:45Time-dependent Increase in the Network Response to the Stimulation of Neuronal Cell Cultures on Micro-electrode Arrays
Published on: May 29, 2017
07:14Tracking Infiltration Front Depth Using Time-lapse Multi-offset Gathers Collected with Array Antenna Ground Penetrating Radar
Published on: May 1, 2018
相关概念视频
State Space Representation
Consider an RLC circuit, a...
Echo
Imagine the sound is reflected back to the ears. Assuming that the source is very close to the human, the difference between hearing the two sounds—the emitted sound and the reflected sound—may be more than the minimum time for perceiving distinct sounds. If this is the case,...
Classification of Systems-II
Transient and Steady-state Response
These test signals are integral in designing control systems to exhibit two key performance aspects: transient response and steady-state...
Time-Domain Interpretation of PD Control
Consider the example of control of motor torque. Initially, a positive...
First Order Systems
When a first-order system is subjected to a unit-step input, its response is characterized by its transfer function. By applying the Laplace transform of the unit-step input to the transfer function, expanding the...
