Search results for: neural network approximation

Authors: Wajdi Bellil, Chokri Ben Amar, Adel M. Alimi

Abstract:

This paper proposes a comparison between wavelet neural networks (WNN), RBF neural network and polynomial approximation in term of 1-D and 2-D functions approximation. We present a novel wavelet neural network, based on Beta wavelets, for 1-D and 2-D functions approximation. Our purpose is to approximate an unknown function f: Rn - R from scattered samples (xi; y = f(xi)) i=1....n, where first, we have little a priori knowledge on the unknown function f: it lives in some infinite dimensional smooth function space and second the function approximation process is performed iteratively: each new measure on the function (xi; f(xi)) is used to compute a new estimate f as an approximation of the function f. Simulation results are demonstrated to validate the generalization ability and efficiency of the proposed Beta wavelet network.

Keywords: Beta wavelets networks, RBF neural network, training algorithms, MSE, 1-D, 2D function approximation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1918

Authors: A. Jabbari, R. Jedermann, W. Lang

Abstract:

The new idea of this research is application of a new fault detection and isolation (FDI) technique for supervision of sensor networks in transportation system. In measurement systems, it is necessary to detect all types of faults and failures, based on predefined algorithm. Last improvements in artificial neural network studies (ANN) led to using them for some FDI purposes. In this paper, application of new probabilistic neural network features for data approximation and data classification are considered for plausibility check in temperature measurement. For this purpose, two-phase FDI mechanism was considered for residual generation and evaluation.

Keywords: Fault detection and Isolation, Neural network, Temperature measurement, measurement approximation and classification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2069

Authors: Wenbin Zhang, Hantian Wu, Jian Tang

Abstract:

An artificial neural network is a mathematical model inspired by biological neural networks. There are several kinds of neural networks and they are widely used in many areas, such as: prediction, detection, and classification. Meanwhile, in day to day life, people always have to make many difficult decisions. For example, the coach of a soccer club has to decide which offensive player to be selected to play in a certain game. This work describes a novel Neural Network using a combination of the General Regression Neural Network and the Probabilistic Neural Networks to help a soccer coach make an informed decision.

Keywords: General Regression Neural Network, Probabilistic Neural Networks, Neural function.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3763

Authors: Mohammad Reza Ghasemi, Amin Ghorbani

Abstract:

The main goal of the present work is to decrease the computational burden for optimum design of steel frames with frequency constraints using a new type of neural networks called Wavelet Neural Network. It is contested to train a suitable neural network for frequency approximation work as the analysis program. The combination of wavelet theory and Neural Networks (NN) has lead to the development of wavelet neural networks. Wavelet neural networks are feed-forward networks using wavelet as activation function. Wavelets are mathematical functions within suitable inner parameters, which help them to approximate arbitrary functions. WNN was used to predict the frequency of the structures. In WNN a RAtional function with Second order Poles (RASP) wavelet was used as a transfer function. It is shown that the convergence speed was faster than other neural networks. Also comparisons of WNN with the embedded Artificial Neural Network (ANN) and with approximate techniques and also with analytical solutions are available in the literature.

Keywords: Weight Minimization, Frequency Constraints, Steel Frames, ANN, WNN, RASP Function.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1738

Authors: Tienfuan Kerh, Hsienchang Lu, Rob Saunders

Abstract:

Shoreline erosion problems caused by global warming and sea level rising may result in losing of land areas, so it should be examined regularly to reduce possible negative impacts. Initially in this study, three sets of survey images obtained from the years of 1990, 2001, and 2010, respectively, are digitalized by using graphical software to establish the spatial coordinates of six major beaches around the island of Taiwan. Then, by overlaying the known multi-period images, the change of shoreline can be observed from their distribution of coordinates. In addition, the neural network approximation is used to develop a model for predicting shoreline variation in the years of 2015 and 2020. The comparison results show that there is no significant change of total sandy area for all beaches in the three different periods. However, the prediction results show that two beaches may exhibit an increasing of total sandy areas under a statistical 95% confidence interval. The proposed method adopted in this study may be applicable to other shorelines of interest around the world.

Keywords: Digitalized shoreline coordinates, survey image overlaying, neural network approximation, total beach sandy areas.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2015

Authors: Wajdi Bellil, Chokri Ben Amar, Adel M. Alimi

Abstract:

An application of Beta wavelet networks to synthesize pass-high and pass-low wavelet filters is investigated in this work. A Beta wavelet network is constructed using a parametric function called Beta function in order to resolve some nonlinear approximation problem. We combine the filter design theory with wavelet network approximation to synthesize perfect filter reconstruction. The order filter is given by the number of neurons in the hidden layer of the neural network. In this paper we use only the first derivative of Beta function to illustrate the proposed design procedures and exhibit its performance.

Keywords: Beta wavelets, Wavenet, multiresolution analysis, perfect filter reconstruction, salient point detect, repeatability.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1663

Authors: Ashok Kumar Goel, Suresh Chandra Saxena, Surekha Bhanot

Abstract:

In this work, a Modified Functional Link Artificial Neural Network (M-FLANN) is proposed which is simpler than a Multilayer Perceptron (MLP) and improves upon the universal approximation capability of Functional Link Artificial Neural Network (FLANN). MLP and its variants: Direct Linear Feedthrough Artificial Neural Network (DLFANN), FLANN and M-FLANN have been implemented to model a simulated Water Bath System and a Continually Stirred Tank Heater (CSTH). Their convergence speed and generalization ability have been compared. The networks have been tested for their interpolation and extrapolation capability using noise-free and noisy data. The results show that M-FLANN which is computationally cheap, performs better and has greater generalization ability than other networks considered in the work.

Keywords: DLFANN, FLANN, M-FLANN, MLP

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1802

Authors: V Krishnaveni, S Jayaraman, A Gunasekaran, K Ramadoss

Abstract:

The ElectroEncephaloGram (EEG) is useful for clinical diagnosis and biomedical research. EEG signals often contain strong ElectroOculoGram (EOG) artifacts produced by eye movements and eye blinks especially in EEG recorded from frontal channels. These artifacts obscure the underlying brain activity, making its visual or automated inspection difficult. The goal of ocular artifact removal is to remove ocular artifacts from the recorded EEG, leaving the underlying background signals due to brain activity. In recent times, Independent Component Analysis (ICA) algorithms have demonstrated superior potential in obtaining the least dependent source components. In this paper, the independent components are obtained by using the JADE algorithm (best separating algorithm) and are classified into either artifact component or neural component. Neural Network is used for the classification of the obtained independent components. Neural Network requires input features that exactly represent the true character of the input signals so that the neural network could classify the signals based on those key characters that differentiate between various signals. In this work, Auto Regressive (AR) coefficients are used as the input features for classification. Two neural network approaches are used to learn classification rules from EEG data. First, a Polynomial Neural Network (PNN) trained by GMDH (Group Method of Data Handling) algorithm is used and secondly, feed-forward neural network classifier trained by a standard back-propagation algorithm is used for classification and the results show that JADE-FNN performs better than JADEPNN.

Keywords: Auto Regressive (AR) Coefficients, Feed Forward Neural Network (FNN), Joint Approximation Diagonalisation of Eigen matrices (JADE) Algorithm, Polynomial Neural Network (PNN).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1888

Authors: Chaghoub Soraya, Zhang Xiaoyan

Abstract:

This research presents the first constant approximation algorithm to the p-median network design problem with multiple cable types. This problem was addressed with a single cable type and there is a bifactor approximation algorithm for the problem. To the best of our knowledge, the algorithm proposed in this paper is the first constant approximation algorithm for the p-median network design with multiple cable types. The addressed problem is a combination of two well studied problems which are p-median problem and network design problem. The introduced algorithm is a random sampling approximation algorithm of constant factor which is conceived by using some random sampling techniques form the literature. It is based on a redistribution Lemma from the literature and a steiner tree problem as a subproblem. This algorithm is simple, and it relies on the notions of random sampling and probability. The proposed approach gives an approximation solution with one constant ratio without violating any of the constraints, in contrast to the one proposed in the literature. This paper provides a (21 + 2)-approximation algorithm for the p-median network design problem with multiple cable types using random sampling techniques.

Keywords: Approximation algorithms, buy-at-bulk, combinatorial optimization, network design, p-median.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 595

Authors: Mohammad Heidari

Abstract:

Horizontal wells are proven to be better producers because they can be extended for a long distance in the pay zone. Engineers have the technical means to forecast the well productivity for a given horizontal length. However, experiences have shown that the actual production rate is often significantly less than that of forecasted. It is a difficult task, if not impossible to identify the real reason why a horizontal well is not producing what was forecasted. Often the source of problem lies in the drilling of horizontal section such as permeability reduction in the pay zone due to mud invasion or snaky well patterns created during drilling. Although drillers aim to drill a constant inclination hole in the pay zone, the more frequent outcome is a sinusoidal wellbore trajectory. The two factors, which play an important role in wellbore tortuosity, are the inclination and side force at bit. A constant inclination horizontal well can only be drilled if the bit face is maintained perpendicular to longitudinal axis of bottom hole assembly (BHA) while keeping the side force nil at the bit. This approach assumes that there exists no formation force at bit. Hence, an appropriate BHA can be designed if bit side force and bit tilt are determined accurately. The Artificial Neural Network (ANN) is superior to existing analytical techniques. In this study, the neural networks have been employed as a general approximation tool for estimation of the bit side forces. A number of samples are analyzed with ANN for parameters of bit side force and the results are compared with exact analysis. Back Propagation Neural network (BPN) is used to approximation of bit side forces. Resultant low relative error value of the test indicates the usability of the BPN in this area.

Keywords: Artificial Neural Network, BHA, Horizontal Well, Stabilizer.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1977

Authors: S. Janpong, K-L. Areerak, K-N. Areerak

Abstract:

This paper aims to present the reviews of the application of neural network in shunt active power filter (SAPF). From the review, three out of four components of SAPF structure, which are harmonic detection component, compensating current control, and DC bus voltage control, have been adopted some of neural network architecture as part of its component or even substitution. The objectives of most papers in using neural network in SAPF are to increase the efficiency, stability, accuracy, robustness, tracking ability of the systems of each component. Moreover, minimizing unneeded signal due to the distortion is the ultimate goal in applying neural network to the SAPF. The most famous architecture of neural network in SAPF applications are ADALINE and Backpropagation (BP).

Keywords: Active power filter, neural network, harmonic distortion, harmonic detection and compensation, non-linear load.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3065

Authors: B. Chandra, P. Paul Varghese

Abstract:

Crypto System Identification is one of the challenging tasks in Crypt analysis. The paper discusses the possibility of employing Neural Networks for identification of Cipher Systems from cipher texts. Cascade Correlation Neural Network and Back Propagation Network have been employed for identification of Cipher Systems. Very large collection of cipher texts were generated using a Block Cipher (Enhanced RC6) and a Stream Cipher (SEAL). Promising results were obtained in terms of accuracy using both the Neural Network models but it was observed that the Cascade Correlation Neural Network Model performed better compared to Back Propagation Network.

Keywords: Back Propagation Neural Networks, CascadeCorrelation Neural Network, Crypto systems, Block Cipher, StreamCipher.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2443

Authors: Lina Yan, Jingjing Di, Ke Wang

Abstract:

A new basis function neural network algorithm is proposed for numerical integration. The main idea is to construct neural network model based on spline basis functions, which is used to approximate the integrand by training neural network weights. The convergence theorem of the neural network algorithm, the theorem for numerical integration and one corollary are presented and proved. The numerical examples, compared with other methods, show that the algorithm is effective and has the characteristics such as high precision and the integrand not required known. Thus, the algorithm presented in this paper can be widely applied in many engineering fields.

Keywords: Numerical integration, Spline basis function, Neural network algorithm

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2928

Authors: A. Nasiri Pour, B. Rostami Tabar, A.Rahimzadeh

Abstract:

Accurate demand forecasting is one of the most key issues in inventory management of spare parts. The problem of modeling future consumption becomes especially difficult for lumpy patterns, which characterized by intervals in which there is no demand and, periods with actual demand occurrences with large variation in demand levels. However, many of the forecasting methods may perform poorly when demand for an item is lumpy. In this study based on the characteristic of lumpy demand patterns of spare parts a hybrid forecasting approach has been developed, which use a multi-layered perceptron neural network and a traditional recursive method for forecasting future demands. In the described approach the multi-layered perceptron are adapted to forecast occurrences of non-zero demands, and then a conventional recursive method is used to estimate the quantity of non-zero demands. In order to evaluate the performance of the proposed approach, their forecasts were compared to those obtained by using Syntetos & Boylan approximation, recently employed multi-layered perceptron neural network, generalized regression neural network and elman recurrent neural network in this area. The models were applied to forecast future demand of spare parts of Arak Petrochemical Company in Iran, using 30 types of real data sets. The results indicate that the forecasts obtained by using our proposed mode are superior to those obtained by using other methods.

Keywords: Lumpy Demand, Neural Network, Forecasting, Hybrid Approach.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2679

Authors: M. Ghalambaz, A.R. Noghrehabadi, M.A. Behrang, E. Assareh, A. Ghanbarzadeh, N.Hedayat

Abstract:

This study presents a hybrid neural network and Gravitational Search Algorithm (HNGSA) method to solve well known Wessinger's equation. To aim this purpose, gravitational search algorithm (GSA) technique is applied to train a multi-layer perceptron neural network, which is used as approximation solution of the Wessinger's equation. A trial solution of the differential equation is written as sum of two parts. The first part satisfies the initial/ boundary conditions and does not contain any adjustable parameters and the second part which is constructed so as not to affect the initial/boundary conditions. The second part involves adjustable parameters (the weights and biases) for a multi-layer perceptron neural network. In order to demonstrate the presented method, the obtained results of the proposed method are compared with some known numerical methods. The given results show that presented method can introduce a closer form to the analytic solution than other numerical methods. Present method can be easily extended to solve a wide range of problems.

Keywords: Neural Networks, Gravitational Search Algorithm (GSR), Wessinger's Equation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2397

Authors: Mousavian, M. Moghimi Mofrad, M. H. Vakili, D. Ashouri, R. Alizadeh

Abstract:

The aim of this research is to use artificial neural networks computing technology for estimating the net heating value (NHV) of crude oil by its Properties. The approach is based on training the neural network simulator uses back-propagation as the learning algorithm for a predefined range of analytically generated well test response. The network with 8 neurons in one hidden layer was selected and prediction of this network has been good agreement with experimental data.

Keywords: Neural Network, Net Heating Value, Crude Oil, Experimental, Modeling.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1588

Authors: Motonobu Hattori

Abstract:

In neural networks, when new patterns are learned by a network, the new information radically interferes with previously stored patterns. This drawback is called catastrophic forgetting or catastrophic interference. In this paper, we propose a biologically inspired neural network model which overcomes this problem. The proposed model consists of two distinct networks: one is a Hopfield type of chaotic associative memory and the other is a multilayer neural network. We consider that these networks correspond to the hippocampus and the neocortex of the brain, respectively. Information given is firstly stored in the hippocampal network with fast learning algorithm. Then the stored information is recalled by chaotic behavior of each neuron in the hippocampal network. Finally, it is consolidated in the neocortical network by using pseudopatterns. Computer simulation results show that the proposed model has much better ability to avoid catastrophic forgetting in comparison with conventional models.

Keywords: catastrophic forgetting, chaotic neural network, complementary learning systems, dual-network

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2101

Authors: Khaing Win Mar, Thinn Thu Naing

Abstract:

Nowadays, precipitation prediction is required for proper planning and management of water resources. Prediction with neural network models has received increasing interest in various research and application domains. However, it is difficult to determine the best neural network architecture for prediction since it is not immediately obvious how many input or hidden nodes are used in the model. In this paper, neural network model is used as a forecasting tool. The major aim is to evaluate a suitable neural network model for monthly precipitation mapping of Myanmar. Using 3-layerd neural network models, 100 cases are tested by changing the number of input and hidden nodes from 1 to 10 nodes, respectively, and only one outputnode used. The optimum model with the suitable number of nodes is selected in accordance with the minimum forecast error. In measuring network performance using Root Mean Square Error (RMSE), experimental results significantly show that 3 inputs-10 hiddens-1 output architecture model gives the best prediction result for monthly precipitation in Myanmar.

Keywords: Precipitation prediction, monthly precipitation, neural network models, Myanmar.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1747

Authors: Isao Taguchi, Yasuo Sugai

Abstract:

This paper proposes an efficient learning method for the layered neural networks based on the selection of training data and input characteristics of an output layer unit. Comparing to recent neural networks; pulse neural networks, quantum neuro computation, etc, the multilayer network is widely used due to its simple structure. When learning objects are complicated, the problems, such as unsuccessful learning or a significant time required in learning, remain unsolved. Focusing on the input data during the learning stage, we undertook an experiment to identify the data that makes large errors and interferes with the learning process. Our method devides the learning process into several stages. In general, input characteristics to an output layer unit show oscillation during learning process for complicated problems. The multi-stage learning method proposes by the authors for the function approximation problems of classifying learning data in a phased manner, focusing on their learnabilities prior to learning in the multi layered neural network, and demonstrates validity of the multi-stage learning method. Specifically, this paper verifies by computer experiments that both of learning accuracy and learning time are improved of the BP method as a learning rule of the multi-stage learning method. In learning, oscillatory phenomena of a learning curve serve an important role in learning performance. The authors also discuss the occurrence mechanisms of oscillatory phenomena in learning. Furthermore, the authors discuss the reasons that errors of some data remain large value even after learning, observing behaviors during learning.

Keywords: data selection, function approximation problem, multistage leaning, neural network, voluntary oscillation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1430

Authors: Kelvin Rozier, Vladimir E. Bondarenko

Abstract:

It is known that an analog Hopfield neural network with time delay can generate the outputs which are similar to the human electroencephalogram. To gain deeper insights into the mechanisms of rhythm generation by the Hopfield neural networks and to study the effects of noise on their activities, we investigated the behaviors of the networks with symmetric and asymmetric interneuron connections. The neural network under the study consists of 10 identical neurons. For symmetric (fully connected) networks all interneuron connections aij = +1; the interneuron connections for asymmetric networks form an upper triangular matrix with non-zero entries aij = +1. The behavior of the network is described by 10 differential equations, which are solved numerically. The results of simulations demonstrate some remarkable properties of a Hopfield neural network, such as linear growth of outputs, dependence of synchronization properties on the connection type, huge amplification of oscillation by the external uniform noise, and the capability of the neural network to transform one type of noise to another.

Keywords: Chaos, Hopfield neural network, noise, synchronization

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1890

Authors: Areej Babiker Idris Babiker, Rosdiazli Ibrahim

Abstract:

Gas chromatography (GC) is the most widely used technique in analytical chemistry. However, GC has high initial cost and requires frequent maintenance. This paper examines the feasibility and potential of using a neural network model as an alternative whenever GC is unvailable. It can also be part of system verification on the performance of GC for preventive maintenance activities. It shows the performance of MultiLayer Perceptron (MLP) with Backpropagation structure. Results demonstrate that neural network model when trained using this structure provides an adequate result and is suitable for this purpose. cm.

Keywords: Analyzer, Levenberg-Marquardt, Gas chromatography, Neural network

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1766

Authors: Abderrahmane Amrouche, Jean Michel Rouvaen

Abstract:

In this paper we present an efficient system for independent speaker speech recognition based on neural network approach. The proposed architecture comprises two phases: a preprocessing phase which consists in segmental normalization and features extraction and a classification phase which uses neural networks based on nonparametric density estimation namely the general regression neural network (GRNN). The relative performances of the proposed model are compared to the similar recognition systems based on the Multilayer Perceptron (MLP), the Recurrent Neural Network (RNN) and the well known Discrete Hidden Markov Model (HMM-VQ) that we have achieved also. Experimental results obtained with Arabic digits have shown that the use of nonparametric density estimation with an appropriate smoothing factor (spread) improves the generalization power of the neural network. The word error rate (WER) is reduced significantly over the baseline HMM method. GRNN computation is a successful alternative to the other neural network and DHMM.

Keywords: Speech Recognition, General Regression NeuralNetwork, Hidden Markov Model, Recurrent Neural Network, ArabicDigits.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2183

Authors: Saman M. Abdulla, Najla B. Al-Dabagh, Omar Zakaria

Abstract:

The aim of this article is to explain how features of attacks could be extracted from the packets. It also explains how vectors could be built and then applied to the input of any analysis stage. For analyzing, the work deploys the Feedforward-Back propagation neural network to act as misuse intrusion detection system. It uses ten types if attacks as example for training and testing the neural network. It explains how the packets are analyzed to extract features. The work shows how selecting the right features, building correct vectors and how correct identification of the training methods with nodes- number in hidden layer of any neural network affecting the accuracy of system. In addition, the work shows how to get values of optimal weights and use them to initialize the Artificial Neural Network.

Keywords: Artificial Neural Network, Attack Features, MisuseIntrusion Detection System, Training Parameters.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2281

Authors: Anupama Pande, Vishik Goel

Abstract:

A complex valued neural network is a neural network, which consists of complex valued input and/or weights and/or thresholds and/or activation functions. Complex-valued neural networks have been widening the scope of applications not only in electronics and informatics, but also in social systems. One of the most important applications of the complex valued neural network is in image and vision processing. In Neural networks, radial basis functions are often used for interpolation in multidimensional space. A Radial Basis function is a function, which has built into it a distance criterion with respect to a centre. Radial basis functions have often been applied in the area of neural networks where they may be used as a replacement for the sigmoid hidden layer transfer characteristic in multi-layer perceptron. This paper aims to present exhaustive results of using RBF units in a complex-valued neural network model that uses the back-propagation algorithm (called 'Complex-BP') for learning. Our experiments results demonstrate the effectiveness of a Radial basis function in a complex valued neural network in image recognition over a real valued neural network. We have studied and stated various observations like effect of learning rates, ranges of the initial weights randomly selected, error functions used and number of iterations for the convergence of error on a neural network model with RBF units. Some inherent properties of this complex back propagation algorithm are also studied and discussed.

Keywords: Complex valued neural network, Radial BasisFunction, Image recognition.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2410

Authors: Defu Zhang, Qingshan Jiang, Xin Li

Abstract:

This paper deals with the application of a well-known neural network technique, multilayer back-propagation (BP) neural network, in financial data mining. A modified neural network forecasting model is presented, and an intelligent mining system is developed. The system can forecast the buying and selling signs according to the prediction of future trends to stock market, and provide decision-making for stock investors. The simulation result of seven years to Shanghai Composite Index shows that the return achieved by this mining system is about three times as large as that achieved by the buy and hold strategy, so it is advantageous to apply neural networks to forecast financial time series, the different investors could benefit from it.

Keywords: Data mining, neural network, stock forecasting.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3589

Authors: Anupama Pande, Ashok Kumar Thakur, Swapnoneel Roy

Abstract:

A complex valued neural network is a neural network which consists of complex valued input and/or weights and/or thresholds and/or activation functions. Complex-valued neural networks have been widening the scope of applications not only in electronics and informatics, but also in social systems. One of the most important applications of the complex valued neural network is in signal processing. In Neural networks, generalized mean neuron model (GMN) is often discussed and studied. The GMN includes a new aggregation function based on the concept of generalized mean of all the inputs to the neuron. This paper aims to present exhaustive results of using Generalized Mean Neuron model in a complex-valued neural network model that uses the back-propagation algorithm (called -Complex-BP-) for learning. Our experiments results demonstrate the effectiveness of a Generalized Mean Neuron Model in a complex plane for signal processing over a real valued neural network. We have studied and stated various observations like effect of learning rates, ranges of the initial weights randomly selected, error functions used and number of iterations for the convergence of error required on a Generalized Mean neural network model. Some inherent properties of this complex back propagation algorithm are also studied and discussed.

Keywords: Complex valued neural network, Generalized Meanneuron model, Signal processing.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1729

Authors: Jiann-Ming Wu, Hsiao-Chang Chen, Chun-Chang Wu, Pei-Hsun Hsu

Abstract:

This work explores blind image deconvolution by recursive function approximation based on supervised learning of neural networks, under the assumption that a degraded image is linear convolution of an original source image through a linear shift-invariant (LSI) blurring matrix. Supervised learning of neural networks of radial basis functions (RBF) is employed to construct an embedded recursive function within a blurring image, try to extract non-deterministic component of an original source image, and use them to estimate hyper parameters of a linear image degradation model. Based on the estimated blurring matrix, reconstruction of an original source image from a blurred image is further resolved by an annealed Hopfield neural network. By numerical simulations, the proposed novel method is shown effective for faithful estimation of an unknown blurring matrix and restoration of an original source image.

Keywords: Blind image deconvolution, linear shift-invariant(LSI), linear image degradation model, radial basis functions (rbf), recursive function, annealed Hopfield neural networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2060

Authors: Apoorva Vinod, Archana Mathur, Snehanshu Saha

Abstract:

Though there exists a plethora of Activation Functions (AFs) used in single and multiple hidden layer Neural Networks (NN), their behavior always raised curiosity, whether used in combination or singly. The popular AFs – Sigmoid, ReLU, and Tanh – have performed prominently well for shallow and deep architectures. Most of the time, AFs are used singly in multi-layered NN, and, to the best of our knowledge, their performance is never studied and analyzed deeply when used in combination. In this manuscript, we experiment on multi-layered NN architecture (both on shallow and deep architectures; Convolutional NN and VGG16) and investigate how well the network responds to using two different AFs (Sigmoid-Tanh, Tanh-ReLU, ReLU-Sigmoid) used alternately against a traditional, single (Sigmoid-Sigmoid, Tanh-Tanh, ReLU-ReLU) combination. Our results show that on using two different AFs, the network achieves better accuracy, substantially lower loss, and faster convergence on 4 computer vision (CV) and 15 Non-CV (NCV) datasets. When using different AFs, not only was the accuracy greater by 6-7%, but we also accomplished convergence twice as fast. We present a case study to investigate the probability of networks suffering vanishing and exploding gradients when using two different AFs. Additionally, we theoretically showed that a composition of two or more AFs satisfies Universal Approximation Theorem (UAT).

Keywords: Activation Function, Universal Approximation function, Neural Networks, convergence.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 154

Authors: Thomas Behan, Zaiyi Liao, Lian Zhao, Chunting Yang

Abstract:

In this paper, low end Digital Signal Processors (DSPs) are applied to accelerate integer neural networks. The use of DSPs to accelerate neural networks has been a topic of study for some time, and has demonstrated significant performance improvements. Recently, work has been done on integer only neural networks, which greatly reduces hardware requirements, and thus allows for cheaper hardware implementation. DSPs with Arithmetic Logic Units (ALUs) that support floating or fixed point arithmetic are generally more expensive than their integer only counterparts due to increased circuit complexity. However if the need for floating or fixed point math operation can be removed, then simpler, lower cost DSPs can be used. To achieve this, an integer only neural network is created in this paper, which is then accelerated by using DSP instructions to improve performance.

Keywords: Digital Signal Processor (DSP), Integer Neural Network(INN), Low Cost Neural Network, Integer Neural Network DSPImplementation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1795

Authors: I. Oloyede

Abstract:

The study investigates the accuracy and loss of deep learning algorithms with the set of coronavirus (COVID-19) images dataset by comparing Bayesian convolutional neural network and traditional convolutional neural network in low dimensional dataset. 50 sets of X-ray images out of which 25 were COVID-19 and the remaining 20 were normal, twenty images were set as training while five were set as validation that were used to ascertained the accuracy of the model. The study found out that Bayesian convolution neural network outperformed conventional neural network at low dimensional dataset that could have exhibited under fitting. The study therefore recommended Bayesian Convolutional neural network (BCNN) for android apps in computer vision for image detection.

Keywords: BCNN, CNN, Images, COVID-19, Deep Learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 871