DeePr-ESN: A deep projection-encoding echo-state network

Neuromorphic overparameterisation and few-shot learning in multilayer physical neural networks

Physical neuromorphic computing, exploiting the complex dynamics of physical systems, has seen rapid advancements in sophistication and performance. Physical reservoir computing, a subset of neuromorphic computing, faces limitations due to its reliance on single systems. This constrains output dimensionality and dynamic range, limiting performance to a narrow range of tasks. Here, we engineer a suite of nanomagnetic array physical reservoirs and interconnect them in parallel and series to create a multilayer neural network architecture. The output of one reservoir is recorded, scaled and virtually fed as input to the next reservoir. This networked approach increases output dimensionality, internal dynamics and computational performance. We demonstrate that a physical neuromorphic system can achieve an overparameterised state, facilitating meta-learning on small training sets and yielding strong performance across a wide range of tasks. Our approach's efficacy is further demonstrated through few-shot learning, where the system rapidly adapts to new tasks.

A lightweight method of integrated local load forecasting and control of edge computing in active distribution networks

The strong resource constraints of edge-computing devices and the dynamic evolution of load characteristics put forward higher requirements for forecasting methods of active distribution networks. This paper proposes a lightweight adaptive ensemble learning method for local load forecasting and predictive control of active distribution networks based on edge computing in resource constrained scenarios. First, the adaptive sparse integration method is proposed to reduce the model scale. Then, the auto-encoder is introduced to downscale the model variables to further reduce computation time and storage overhead. An adaptive correction method is proposed to maintain the adaptability. Finally, a multi-timescale predictive control method for the edge side is established, which realizes the collaboration of local load forecasting and control. All cases can be deployed on an actual edge-computing device. Compared to other benchmark methods and the existing researches, the proposed method can minimize the model complexity without reducing the forecasting accuracy.

fastESN: Fast Echo State Network

Echo state networks (ESNs) are reservoir computing-based recurrent neural networks widely used in pattern analysis and machine intelligence applications. In order to achieve high accuracy with large model capacity, ESNs usually contain a large-sized internal layer (reservoir), making the evaluation process too slow for some applications. In this work, we speed up the evaluation of ESN by building a reduced network called the fast ESN (fastESN) and achieve an ESN evaluation complexity independent of the original ESN size for the first time. FastESN is generated using three techniques. First, the high-dimensional state of the original ESN is approximated by a low-dimensional state through proper orthogonal decomposition (POD)-based projection. Second, the activation function evaluation number is reduced through the discrete empirical interpolation method (DEIM). Third, we show the directly generated fastESN has instability problems and provide a stabilization scheme as a solution. Through experiments on four popular benchmarks, we show that fastESN is able to accelerate the sparse storage-based ESN evaluation with a high parameter compression ratio and a fast evaluation speed.

Hierarchical Echo State Network With Sparse Learning: A Method for Multidimensional Chaotic Time Series Prediction

Echo state network (ESN), a type of special recurrent neural network with a large-scale randomly fixed hidden layer (called a reservoir) and an adaptable linear output layer, has been widely employed in the field of time series analysis and modeling. However, when tackling the problem of multidimensional chaotic time series prediction, due to the randomly generated rules for input and reservoir weights, not only the representation of valuable variables is enriched but also redundant and irrelevant information is accumulated inevitably. To remove the redundant components, reduce the approximate collinearity among echo-state information, and improve the generalization and stability, a new method called hierarchical ESN with sparse learning (HESN-SL) is proposed. The HESN-SL mines and captures the latent evolution patterns hidden from the dynamic system by means of layer-by-layer processing in stacked reservoirs, and leverage monotone accelerated proximal gradient algorithm to train a sparse output layer with variable selection capability. Meanwhile, we further prove that the HESN-SL satisfies the echo state property, which guarantees the stability and convergence of the proposed model when applied to time series prediction. Experimental results on two synthetic chaotic systems and a real-world meteorological dataset illustrate the proposed HESN-SL outperforms both original ESN and existing hierarchical ESN-based models for multidimensional chaotic time series prediction.

Chain-Structure Echo State Network With Stochastic Optimization: Methodology and Application

In this article, a chain-structure echo state network (CESN) with stacked subnetwork modules is newly proposed as a new kind of deep recurrent neural network for multivariate time series prediction. Motivated by the philosophy of "divide and conquer," the related input vectors are first divided into clusters, and the final output results of CESN are then integrated by successively learning the predicted values of each clustered variable. Network structure, mathematical model, training mechanism, and stability analysis are, respectively, studied for the proposed CESN. In the training stage, least-squares regression is first used to pretrain the output weights in a module-by-module way, and stochastic local search (SLS) is developed to fine-tune network weights toward global optima. The loss function of CESN can be effectively reduced by SLS. To avoid overfitting, the optimization process is stopped when the validation error starts to increase. Finally, SLS-CESN is evaluated in chaos prediction benchmarks and real applications. Four different examples are given to verify the effectiveness and robustness of CESN and SLS-CESN.

Bayesian Optimized Echo State Network Applied to Short-Term Load Forecasting

Load forecasting impacts directly financial returns and information in electrical systems planning. A promising approach to load forecasting is the Echo State Network (ESN), a recurrent neural network for the processing of temporal dependencies. The low computational cost and powerful performance of ESN make it widely used in a range of applications including forecasting tasks and nonlinear modeling. This paper presents a Bayesian optimization algorithm (BOA) of ESN hyperparameters in load forecasting with its main contributions including helping the selection of optimization algorithms for tuning ESN to solve real-world forecasting problems, as well as the evaluation of the performance of Bayesian optimization with different acquisition function settings. For this purpose, the ESN hyperparameters were set as variables to be optimized. Then, the adopted BOA employs a probabilist model using Gaussian process to find the best set of ESN hyperparameters using three different options of acquisition function and a surrogate utility function. Finally, the optimized hyperparameters are used by the ESN for predictions. Two datasets have been used to test the effectiveness of the proposed forecasting ESN model using BOA approaches, one from Poland and another from Brazil. The results of optimization statistics, convergence curves, execution time profile, and the hyperparameters’ best solution frequencies indicate that each problem requires a different setting for the BOA. Simulation results are promising in terms of short-term load forecasting quality and low error predictions may be achieved, given the correct options settings are used. Furthermore, since there is not an optimal global optimization solution known for real-world problems, correlations among certain values of hyperparameters are useful to guide the selection of such a solution.