Current Issue : January - March Volume : 2019 Issue Number : 1 Articles : 5 Articles
This article is devoted to a time series prediction scheme involving the nonlinear\nautoregressive algorithm and its applications. The scheme is implemented\nby means of an artificial neural network containing a hidden layer. As\na training algorithm we use scaled conjugate gradient (SCG) method and the\nBayesian regularization (BReg) method. The first method is applied to time\nseries without noise, while the second one can also be applied for noisy datasets.\nWe apply the suggested scheme for prediction of time series arising in oil\nand gas pricing using 50 and 100 past values. Results of numerical simulations\nare presented and discussed....
We propose an anomaly detection approach by learning a generative model using deep neural network. A weighted convolutional\nautoencoder- (AE-) long short-termmemory (LSTM) network is proposed to reconstruct rawdata and performanomaly detection\nbased on reconstruction errors to resolve the existing challenges of anomaly detection in complicated definitions and background\ninfluence. Convolutional AEs and LSTMs are used to encode spatial and temporal variations of input frames, respectively. A\nweighted Euclidean loss is proposed to enable the network to concentrate on moving foregrounds, thus restraining background\ninfluence. Moving foregrounds are segmented from the input frames using robust principal component analysis decomposition.\nComparisons with state-of-the-art approaches indicate the superiority of our approach in anomaly detection. Generalization of\nanomaly detection is improved by enforcing the network to focus on moving foregrounds...
An artificial neural network (ann), adaptive neurofuzzy inference system (anfis) models, and fuzzy rule-based system (frbs)\nmodels are developed to predict the attendance demand in European football games, in this paper. To determine the most\nsuccessful method, each of the methods is analyzed under different situations. The Elman backpropagation, feed-forward\nbackpropagation, and cascade-forward backpropagation network types are developed to determine the outperforming ann\nmodel. The backpropagation and hybrid optimization methods are used for training fuzzy inference system (fis) to determine the\noutperforming anfis model. The fuzzy logic model is developed after experimenting different forms of membership functions.\nTo this end, the data of 236 soccer games are used to train the ann and anfis models, and 2017/2018 seasonâ??s data of these clubs\nare used to test all of the models. The results of all models are compared with each other and real past data. To assess the\nperformance of each model, two error measures that are Mean Absolute Percent Error (mape) and Mean Absolute Deviation\n(mad) are implemented. These measures reveal that the ANN model that has Elman network type outperforms the other models.\nFinally, the results emphasize that the proposed ANN model can be effectively used for prediction purposes....
This paper proposes a method for estimating traffic flows on some links of a road network\nknowing the data on other links that are monitored with sensors. In this way, it is possible to\nobtain more information on traffic conditions without increasing the number of monitored links.\nThe proposed method is based on artificial neural networks (ANNs), wherein the input data are\nthe traffic flows on some monitored road links and the output data are the traffic flows on some\nunmonitored links. We have implemented and tested several single-layer feed-forward ANNs that\ndiffer in the number of neurons and the method of generating datasets for training. The proposed\nANNs were trained with a supervised learning approach where input and output example datasets\nwere generated through traffic simulation techniques. The proposed method was tested on a\nreal-scale network and gave very good results if the travel demand patterns were known and used\nfor generating example datasets, and promising results if the demand patterns were not considered\nin the procedure. Numerical results have underlined that the ANNs with few neurons were more\neffective than the ones with many neurons in this specific problem....
This research proposes a genetic algorithm that provides a solution to the problem of\ndeficient distribution of drinking water via the current hydraulic network in the neighborhood\nâ??Fraccionamiento Real Montecasinoâ? (FRM), in Huitzilac, Morelos, Mexico. The proposed solution is\nthe addition of new elements to the FRM network. The new elements include storage tanks, pipes,\nand pressure-reducing valves. To evaluate the constraint satisfaction model of mass and energy\nconservation, the hydraulic EPANET solver (HES) is used with an optimization model to minimize\nthe total cost of changes in the network (new pipes, tanks, and valves). A genetic algorithm was used\nto evaluate the optimization model. The analysis of the results obtained by the genetic algorithm\nfor the FRM network shows that adequate and balanced pressures were obtained by means of small\nmodifications to the existing network, which entailed minimal costs. Simulations were performed\nfor an extended period, which means that the pressure was obtained by simulation with HSE at\none-hour intervals, during the algorithm execution, to verify adequate pressure at a specific point in\nthe system, or to make corrections to ensure proper distribution, this with the aim of having a final\noptimized network design....
Loading....