Performance boost of time-delay reservoir computing by non-resonant clock cycle

  title={Performance boost of time-delay reservoir computing by non-resonant clock cycle},
  author={Florian Stelzer and Andr{\'e} R{\"o}hm and Kathy L{\"u}dge and Serhiy Yanchuk},
  journal={Neural networks : the official journal of the International Neural Network Society},

Figures from this paper

Deep Time-Delay Reservoir Computing: Dynamics and Memory Capacity

This work presents how the dynamical properties of a deep Ikeda-based reservoir are related to its memory capacity (MC) and how that can be used for optimization and presents two configurations that empower either high nonlinear MC or long time linear MC.

The role of delay-times in delay-based Photonic Reservoir Computing

The existing literature on this subject is reviewed and the concept of delay-based reservoir computing is introduced in a manner that demonstrates that there is no predefined relationship between these two times-scales.

Limitations of the recall capabilities in delay based reservoir computing systems

The results suggest that even for constant readout dimension the total memory capacity is dependent on the ratio between the information input period, also called the clock cycle, and the time delay in the system.

Insight into delay based reservoir computing via eigenvalue analysis

This paper applies the method exemplarily to a photonic laser system with feedback and compares the numerically computed recall capabilities with the eigenvalue spectrum, suggesting that any dynamical system used as a reservoir can be analysed in this way.

Reservoir Computing with Delayed Input for Fast and Easy Optimisation

It is demonstrated that by including a time-delayed version of the input for various time series prediction tasks, good performance can be achieved with an unoptimised reservoir and it is shown that one unaltered reservoir can perform well on six different time series Prediction tasks at a very low computational expense.

Master memory function for delay-based reservoir computers with single-variable dynamics

This work shows that many delay-based reservoir computers considered in the literature can be characterized by a universal master memory function (MMF), and proposes an analytical description of the MMF that enables its efficient and fast computation.

Impact of phase modulation on the performance of photonic delay-based reservoir computing with semiconductor lasers

In photonic reservoir computing, semiconductor lasers with delayed feedback have been used to efficiently solve difficult and time-consuming problems. The injection of data in these systems is often

Reservoir Computing with Diverse Timescales for Prediction of Multiscale Dynamics

A reservoir computing model with diverse timescales by using a recurrent network of heterogeneous leaky integrator (LI) neurons is proposed, and it is demonstrated that a closed-loop version of the proposed model can achieve longer-term predictions compared to the counterpart with identical LI neurons depending on the hyperparameter setting.

Deep neural networks using a single neuron: folded-in-time architecture using feedback-modulated delay loops

This work presents a method for folding a deep neural network of arbitrary size into a single neuron with multiple time-delayed feedback loops, which can fully represent standard Deep Neural Networks, encompasses sparse DNNs, and extends the DNN concept toward dynamical systems implementations.

Improving Delay Based Reservoir Computing via Eigenvalue Analysis

It is shown that these two quantities are deeply connected, and thus the reservoir computing performance is predictable by analyzing the eigenvalue spectrum, and any dynamical system used as a reservoir can be analyzed in this way as long as the reservoir perturbations are sufficiently small.



Introduction to Focus Issue: Time-delay dynamics.

Current hot problems such as the synchronization properties of networks of delay-coupled units, novel stabilization techniques, and the large delay limit of a delay differential equation are addressed in this special issue.

Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations

A new computational model for real-time computing on time-varying input that provides an alternative to paradigms based on Turing machines or attractor neural networks, based on principles of high-dimensional dynamical systems in combination with statistical learning theory and can be implemented on generic evolved or found recurrent circuitry.

Tutorial: Photonic Neural Networks in Delay Systems

The most relevant aspects of Artificial Neural Networks and delay systems are introduced, the seminal experimental demonstrations of Reservoir Computing in photonic delay systems, plus the most recent and advanced realizations are explained.

Multiplexed networks: reservoir computing with virtual and real nodes

This work investigates the reservoir computing performance of hybrid network-delay systems systematically by evaluating the NARMA10 and the Sante Fe task for varying system parameters, and constructs ‘multiplexed networks’ that can be seen as intermediate steps on the scale from classical networks to the ‘virtual networks' of delay systems.

Advances in photonic reservoir computing

A novel paradigm that has emerged in analogue neuromorphic optical computing is reviewed: networks implemented with multiple discrete optical nodes and the continuous system of a single nonlinear device coupled to delayed feedback.

Impact of input mask signals on delay-based photonic reservoir computing with semiconductor lasers.

The chaos mask signal shows superior performance than that of the digital mask signals, however, similar prediction errors can be achieved for the chaos and colored-noise mask signals.

An Analytical Approach to Single Node Delay-Coupled Reservoir Computing

This work derives here approximate analytical equations for the reservoir by solving the underlying system explicitly, which represents the system accurately and yields comparable performance in reservoir benchmark tasks, while reducing computational costs by several orders of magnitude.

Information Processing Capacity of Dynamical Systems

The theory combines concepts from machine learning (reservoir computing), system modeling, stochastic processes, and functional analysis to define the computational capacity of a dynamical system.

Optoelectronic Reservoir Computing

This work reports an optoelectronic implementation of reservoir computing based on a recently proposed architecture consisting of a single non linear node and a delay line that is sufficiently fast for real time information processing.

Information processing using a single dynamical node as complex system

This work introduces a novel architecture that reduces the usually required large number of elements to a single nonlinear node with delayed feedback and proves that delay-dynamical systems, even in their simplest manifestation, can perform efficient information processing.