Improved second-order training algorithms for globally and partially recurrent neural networks

Santos, E.P. dos; Zuben, Fernando J. Von

doi:10.1109/ijcnn.1999.832591

Cited by 10 publications

(7 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…43 An additional step of the computational analysis involved the application of the Friedman's test 26 to verify whether the obtained results could be considered significantly different. In all cases, the p-values achieved were close to zero, which indicates that the results were directly affected by the chosen predictor.…”

Section: Resultsmentioning

confidence: 99%

Unorganized Machines for Seasonal Streamflow Series Forecasting

Siqueira

Boccato

Attux

et al. 2014

Int. J. Neur. Syst.

View full text Add to dashboard Cite

Modern unorganized machines--extreme learning machines and echo state networks--provide an elegant balance between processing capability and mathematical simplicity, circumventing the difficulties associated with the conventional training approaches of feedforward/recurrent neural networks (FNNs/RNNs). This work performs a detailed investigation of the applicability of unorganized architectures to the problem of seasonal streamflow series forecasting, considering scenarios associated with four Brazilian hydroelectric plants and four distinct prediction horizons. Experimental results indicate the pertinence of these models to the focused task.

show abstract

Section: Resultsmentioning

confidence: 99%

Unorganized Machines for Seasonal Streamflow Series Forecasting

Siqueira

Boccato

Attux

et al. 2014

Int. J. Neur. Syst.

View full text Add to dashboard Cite

show abstract

“…Here, we use an MLP with three layers, with hyperbolic tangent and linear function as activation functions for the hidden and output layers, respectively. However, in this work the training process is performed using the Modified Scaled Conjugated Gradient [62].…”

Section: ) Multilayer Perceptron -Mlpmentioning

confidence: 99%

“…The MLP is trained via the Modified Scale Conjugated Gradient [62] algorithm using the following stopping conditions: (i) the maximum number of iterations equal 300; (ii) use of the hold-out cross-validation; (iii) progress training 10 6 . The number of hidden neurons for a single MLP model is defined using a grid search in the range [3,250].…”

Section: B Experimental Setupmentioning

confidence: 99%

Neural-Based Ensembles for Particulate Matter Forecasting

et al. 2021

View full text Add to dashboard Cite

The air pollution caused by particulate matter (PM) has become a public health issue due to the risks to human life and the environment. The PM concentration in the air causes haze and affects the lungs and the heart, leading to reduced visibility, allergic reactions, pneumonia, asthma, cardiopulmonary diseases, lung cancer, and even death. In this context, the development of systems for monitoring, forecasting, and controlling emissions plays an important role. The literature about forecasting systems based on Artificial Neural Networks (ANNs) ensembles has been highlighted regarding statistical accuracy and efficiency. In this article, trainable and non-trainable combination methods are used for PM 10 and PM 2.5 (particles with an aerodynamic diameter less than 10 and 2.5 micrometers, respectively) time series forecasting for eight different locations, in Finland and Brazil, for different periods. Trainable ensembles based on ANNs, linear regression, and Copulas are compared with non-trainable combinations (mean and median), single ANNs, and linear statistical approaches. Different models are considered so far, including Autoregressive model (AR), Autoregressive and Moving Average Model (ARMA), Infinite Impulse Response Filters (IIR), Multilayer Perceptron (MLP), Radial Basis Function Networks (RBF), Extreme Learning Machines (ELM), Echo State Networks (ESN), and Adaptive Network Fuzzy Inference System (ANFIS). The use of ANNs ensembles, mainly combined with MLP, leads to a better one step ahead forecasting performance. The use of robust air pollution forecasting tools is prime to assist governments in managing air pollution issues like hospital collapse during adverse air quality situations. In this sense, our study is indirectly related to the following United Nations sustainable development goals: SDG 3 -good health and well-being and SDG 11 -sustainable cities and communities.INDEX TERMS Forecasting, particulate matter, artificial neural networks, ensemble.

show abstract

“…We have adopted here a scaled conjugate gradient second-order method [26], used both for recurrent and nonrecurrent neural networks. This method is numerically stable, has a fast convergence and a low computational cost [27].…”

Section: B Training Processmentioning

confidence: 99%

A Neural Network Controller For The Direct Power Control Of Doubly Fed Induction Generator

Marchi¹,

Dainez²,

Zuben³

et al. 2013

Eletrônica de Potência

View full text Add to dashboard Cite

-In this paper a direct power control strategy for a doubly-fed induction generator by using an artificial neural network controller with the multilayer perceptron structure is presented. The control variables directand quadrature-axis rotor voltage signals are directly generated by proposed controller from both stator current and voltage that are measured by Hall sensors. The input variables of the control system are the rotor speed, the active and reactive power references and their respective errors. The proposed control strategy allows that the converter connected to the rotor terminals operate with constant switching frequency which simplifies the design of the AC harmonic filter and also prevents their power loss. To validate the proposed control strategy, digital simulation and experimental tests are performed for a 2.25 kW doubly-fed induction generator. A TMS320F2812 DSP is used to implement the neural network controller.

show abstract

Improved second-order training algorithms for globally and partially recurrent neural networks

Cited by 10 publications

References 13 publications

Unorganized Machines for Seasonal Streamflow Series Forecasting

Unorganized Machines for Seasonal Streamflow Series Forecasting

Neural-Based Ensembles for Particulate Matter Forecasting

A Neural Network Controller For The Direct Power Control Of Doubly Fed Induction Generator

Contact Info

Product

Resources

About