It has previously been shown that gradient-descent learning algorithms for recurrent neural networks can perform poorly on tasks that involve long-term dependencies, i.e. those problems for which the desired output depends on inputs presented at times far in the past. We show that the long-term dependencies problem is lessened for a class of architectures called nonlinear autoregressive models ...

It has recently been shown that gradient-descent learning algorithms for recurrent neural networks can perform poorly on tasks that involve long-term dependencies, i.e., those problems for which the desired output depends on inputs presented at times far in the past. We show that the long-term dependencies problem is lessened for a class of architectures called Nonlinear AutoRegressive models w...

This paper investigates the technique of the modeling and identification a new dynamic NARX fuzzy model by means of genetic algorithms. In conventional identification techniques, difficulties such as poor knowledge of the process, inaccurate process or complexity of the resulting mathematical model, all which limit their usefulness during dealing with dynamic nonlinear industrial processes. To ...

As renewable energy increasingly integrates into the electric power system, electric load forecasting and renewable energy power generation forecasting become more important. In this project, ARIMA and NARX are applied to build load forecasting model focusing on improving statistical and computational efficiency without losing accuracy. ARIMA turns out to be better for short term forecasting wh...

One of the most powerful and versatile system identification frameworks of the last three decades is the NARMAX/NARX approach, which is based on a nonlinear discrete-time representation. Recent advances in machine learning have motivated new functional forms for the NARX model, including one based on Gaussian processes (GPs), which is the focus of this paper. Because of their nonparametric form...

Abstract Aiming to improve the position and velocity precision of INS/GNSS system during GNSS outages, a novel that combines unscented Kalman filter (UKF) nonlinear autoregressive neural networks with external inputs (NARX) is proposed. The NARX-based module utilized predict measurement updates UKF outages. A new offline approach for selecting optimal NARX suggested tested. This based ...

Complex dynamic behavior of nonlinear structures makes it challenging for uncertainty analysis through Monte Carlo simulations (MCS). Surrogate modeling presents an efficient and accurate computational alternative a large number MCS. The previous study has demonstrated that the multi-input multi-output autoregressive with exogenous input (MIMO-NARX) model provides good discrete-time representat...

Gaussian processes (GP) regression is a powerful probabilistic tool for modeling nonlinear dynamical systems. The downside of the method its cubic computational complexity with respect to training data that can be partially reduced using pseudo-inputs. dynamics represented an autoregressive model, which simplifies static case. When simulating uncertainty propagated through function and simulati...

Due to the advances of network technologies and multimedia communications, Quality of Service (QoS) becomes an increasingly important issue in network communications. Many traditional assessment techniques were designed to evaluate the QoS of multimedia applications transmitted over these networks. In this paper, a new QoS evaluation system has been developed. The proposed system is based on us...

The HAMP linker, a predicted structural element observed in sensor proteins from all domains of life, is proposed to transmit signals between extracellular sensory input domains and cytoplasmic output domains. HAMP (histidine kinase, adenylyl cyclase, methyl-accepting chemotaxis protein, and phosphatase) linkers are located just inside the cytoplasmic membrane and are projected to form two shor...