发表机构
Western University（西安大略大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究提出基于频率的回声状态网络计算，受大脑振荡动力学启发，将其解释为独立振荡单元集合。该方法能理解储层内部行为，性能优于随机储层，可优化改善短期预测，还能预测复杂时空动态，借大脑特性等设计回声状态网络计算。
AI 中文摘要
回声状态网络计算已成为一种预测动力系统生成的时间序列的有效机器学习框架。与其他机器学习和深度学习方法不同，回声状态网络计算仅通过线性回归训练输出层，而不训练储层（循环层）。这种简化使得回声状态网络计算机更易于训练和实验。然而，由于当前的储层由随机连接的节点网络组成，并且需要优化众多超参数，因此仍然缺少一个能够精确解释回声状态网络计算如何工作以及如何优化的框架。在这里，我们提出了一种受大脑振荡动力学及其时间尺度层次结构启发的基于频率的储层。基于频率的储层可以解释为一组独立的振荡单元，每个单元处理输入频率内容的一部分。这使我们能够通过将其建模为由外部输入驱动的单个单元来理解储层的内部行为。借鉴由复杂周期输入驱动的非线性振荡器的理论，我们发现基于频率的储层单元选择性地放大和存储特定的输入频率，然后用于预测。基于频率的储层的性能与等效的随机储层相当或更好。此外，基于频率的方法可以进行优化以改善短期预测，这是随机储层所缺乏的特性。最后，我们表明基于频率的储层还可以预测复杂的时空动态。我们的结果表明，可以利用大脑特性和从强迫非线性振荡器物理学中借鉴的理论见解来设计回声状态网络计算。
英文摘要
Reservoir computing has emerged as an efficient machine learning framework for predicting time series generated by dynamical systems. In contrast to other machine and deep learning approaches, a reservoir computing trains only the output layer via linear regression, leaving the reservoir (recurrent layer) untrained. This simplification makes reservoir computers easier to train and more amenable to experimentation. However, because current reservoirs consist of networks of randomly connected nodes and require the optimization of numerous hyperparameters, a framework that precisely explains how reservoir computing operates and how it can be optimized remains missing. Here, we propose a frequency-based reservoir inspired by the brain's oscillatory dynamics and its hierarchy of timescales. The frequency-based reservoir can be interpreted as an ensemble of independent oscillatory units, each processing a portion of the input's frequency content. This allows us to understand the reservoir's internal behavior by modeling it as a single unit driven by an external input. Borrowing from the theory of a nonlinear oscillator forced by complex periodic inputs, we found that units of the frequency-based reservoir selectively amplify and store specific input frequencies, which are then used for prediction. The frequency-based reservoir performs as well as or better than equivalent random reservoirs. Furthermore, the frequency-based approach can be optimized to improve short-term prediction, a property that random reservoirs lack. Finally, we show that the frequency-based reservoir can also predict complex spatiotemporal dynamics. Our results show that reservoir computing can be designed using brain properties and theoretical insights borrowed from the physics of forced nonlinear oscillators.
Comments24 pages, 8 figures
