Article

Reservoir Computing: uma Abordagem Conceitual

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Reservoir computing é um paradigma de rede neural recorrente construída de forma aleatória, onde sua camada intermediária não necessita ser treinada. O presente artigo sintetiza os principais conceitos, métodos e pesquisas recentes realizadas sobre o paradigma de reservoir computing, objetivando servir como apoio teórico para outros artigos. Foi realizada uma revisão bibliográfica fundamentada em bases de conhecimento científico confiáveis enfatizando pesquisas compreendidas no período de 2007 a 2017 e direcionadas à implementação e otimização do paradigma em questão. Como resultado do trabalho, tem-se a apresentação de trabalhos recentes que contribuem de forma geral para o desenvolvimento de reservoir computing, e devido à atualidade do tema, é apresentada uma diversidade de tópicos abertos à pesquisa, podendo servir como norteamento para a comunidade científica. Palavras-chave: Aprendizado de Máquina. Inteligência Artificial. Redes Neurais Recorrentes.Abstract Reservoir computng is a randomly constructed recurrent neural network paradigm, where the hidden layer does not need to be trained. This article summarizes the main concepts, methods and recent researches about reservoir computing paradigm, aiming to offer a theoretical support for other articles. Were made a bibliographic review based on reliable scientific knowledge bases, emphasizing researches published between 2007 and 2017 and focused on implementation and optimization of aforementioned paradigm. As a result, there's a report of recent articles that contribute in general to the development of reservoir computing, and due to its topicality, a diversity of topics that are still open to research are given, that may possibly work as a guide for the research community. Keywords: Artificial Intelligence. Machine Learning. Recurrent Neural Network.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

ResearchGate has not been able to resolve any citations for this publication.
Thesis
Full-text available
O presente trabalho propõe resolver o clássico problema combinatorial conhecido como Problema do Caixeiro Viajante. Foi usado no sistema de otimização de busca do menor caminho uma rede neural recorrente. A topologia de estrutura de ligação das realimentações da rede adotada aqui é conhecida por Rede Recorrente de Wang. Como regra de treinamento de seus pesos sinápticos foi adotada a técnica de Perturbação Simultânea com Aproximação Estocástica. Foi elaborado ainda uma minuciosa revisão bibliográfica sobre todos os temas abordados com detalhes sobre a otimização multivariável com perturbação simultânea. Comparar-se-á também os resultados obtidos aqui com outras diferentes técnicas aplicadas no Problema do Caixeiro Viajante visando propósitos de validação.
Article
Full-text available
Optimizing a neural network’s topology is a difficult problem for at least two reasons: the topology space is discrete, and the quality of any given topology must be assessed by assigning many different sets of weights to its connections. These two characteristics tend to cause very “rough.” objective functions. Here we demonstrate how self-assembly (SA) and particle swarm optimization (PSO) can be integrated to provide a novel and effective means of concurrently optimizing a neural network’s weights and topology. Combining SA and PSO addresses two key challenges. First, it creates a more integrated representation of neural network weights and topology so that we have just a single, continuous search domain that permits “smoother” objective functions. Second, it extends the traditional focus of self-assembly, from the growth of predefined target structures , to functional self-assembly, in which growth is driven by optimality criteria defined in terms of the performance of emerging structures on predefined computational problems . Our model incorporates a new way of viewing PSO that involves a population of growing, interacting networks, as opposed to particles. The effectiveness of our method for optimizing echo state network weights and topologies is demonstrated through its performance on a number of challenging benchmark problems.
Article
Full-text available
For predicting the key technology indicators (concentrate grade and tailings recovery rate) of flotation process, an echo state network (ESN) based fusion soft-sensor model optimized by the improved glowworm swarm optimization (GSO) algorithm is proposed. Firstly, the color feature (saturation and brightness) and texture features (angular second moment, sum entropy, inertia moment, etc.) based on grey-level co-occurrence matrix (GLCM) are adopted to describe the visual characteristics of the flotation froth image. Then the kernel principal component analysis (KPCA) method is used to reduce the dimensionality of the high-dimensional input vector composed by the flotation froth image characteristics and process datum and extracts the nonlinear principal components in order to reduce the ESN dimension and network complex. The ESN soft-sensor model of flotation process is optimized by the GSO algorithm with congestion factor. Simulation results show that the model has better generalization and prediction accuracy to meet the online soft-sensor requirements of the real-time control in the flotation process.
Article
Full-text available
In this paper, we present an introduction and critical experimental evaluation of a reservoir computing (RC) approach for ambient assisted living (AAL) applications. Such an empirical analysis jointly addresses the issues of efficiency, by analyzing different system configurations toward the embedding into computationally constrained wireless sensor devices, and of efficacy, by analyzing the predictive performance on real-world applications. First, the approach is assessed on a validation scheme where training, validation and test data are sampled in homogeneous ambient conditions, i.e., from the same set of rooms. Then, it is introduced an external test set involving a new setting, i.e., a novel ambient, which was not available in the first phase of model training and validation. The specific test-bed considered in the paper allows us to investigate the capability of the RC approach to discriminate among user movement trajectories from received signal strength indicator sensor signals. This capability can be exploited in various AAL applications targeted at learning user indoor habits, such as in the proposed indoor movement forecasting task. Such a joint analysis of the efficiency/efficacy trade-off provides novel insight in the concrete successful exploitation of RC for AAL tasks and for their distributed implementation into wireless sensor networks.
Article
Full-text available
Electricity demand forecasts are required by companies who need to predict their customers’ demand, and by those wishing to trade electricity as a commodity on financial markets. It is hard to find the right prediction method for a given application if not a prediction expert. Recent works show that Liquid State Machines (LSM’s) can be applied to the prediction of time series. The main advantage of the LSM is that it projects the input data in a high-dimensional dynamical space and therefore simple learning methods can be used to train the readout. In this paper we present an experimental investigation of an approach for the computation of time series prediction by employing Liquid State Machines (LSM) in the modeling of a predictor in a case study for short-term and long-term electricity demand forecasting. Results of this investigation are promising, considering the error to stop training the readout, the number of iterations of training of the readout and that no strategy of seasonal adjustment or preprocessing of data was achieved to extract non-correlated data out of the time series.
Article
Oger (OrGanic Environment for Reservoir computing) is a Python toolbox for building, training and evaluating modular learning architectures on large data sets. It builds on MDP for its modularity, and adds processing of sequential data sets, gradient descent training, several crossvalidation schemes and parallel parameter optimization methods. Additionally, several learning algorithms are implemented, such as different reservoir implementations (both sigmoid and spiking), ridge regression, conditional restricted Boltzmann machine (CRBM) and others, including GPU accelerated versions. Oger is released under the GNU LGPL, and is available from http: //organic.elis.ugent.be/oger.
Article
Three different uses of a recurrent neural network (RNN) as a reservoir that is not trained but instead read out by a simple external classification layer have been described in the literature: Liquid State Machines (LSMs), Echo State Networks (ESNs) and the Backpropagation Decorrelation (BPDC) learning rule. Individual descriptions of these techniques exist, but a overview is still lacking. Here, we present a series of experimental results that compares all three implementations, and draw conclusions about the relation between a broad range of reservoir parameters and network dynamics, memory, node complexity and performance on a variety of benchmark tests with different characteristics. Next, we introduce a new measure for the reservoir dynamics based on Lyapunov exponents. Unlike previous measures in the literature, this measure is dependent on the dynamics of the reservoir in response to the inputs, and in the cases we tried, it indicates an optimal value for the global scaling of the weight matrix, irrespective of the standard measures. We also describe the Reservoir Computing Toolbox that was used for these experiments, which implements all the types of Reservoir Computing and allows the easy simulation of a wide range of reservoir topologies for a number of benchmarks.
Artificial Neural Networks and Machine Learning -ICANN 2012
  • T B Ludermir
  • A T Sergio
LUDERMIR, T.B.; SERGIO, A.T. Artificial Neural Networks and Machine Learning -ICANN 2012. Lecture Notes Computer Sci., v.7552, p.685-692, 2012.
Reservoir Computing: uma Abordagem Conceitual networks for noisy image recognition
Reservoir Computing: uma Abordagem Conceitual networks for noisy image recognition. Neurocomputing, 2017.
Um método para design e treinamento de reservoir computing aplicado à previsão de séries temporais
  • A A Ferreira
FERREIRA, A.A. Um método para design e treinamento de reservoir computing aplicado à previsão de séries temporais. Recife: Universidade Federal de Pernambuco, 2011.
The "echo state" approach to analysing and training recurrent neural networks. no 148. GMD -German National Research Institute for Computer Science
  • H Jaeger
JAEGER. H. The "echo state" approach to analysing and training recurrent neural networks. no 148. GMD -German National Research Institute for Computer Science, 2001. Disponível em: <http://www.faculty.jacobs-university.de/hjaeger/pubs/ EchoStatesTechRep.pdf>. Acesso em 2 out 2017.
On the application of reservoir computing
  • A Jalalvand
JALALVAND, A. et al. On the application of reservoir computing