Dynamic Artificial Neural Network Research Articles

Discriminating lysosomal membrane protein types using dynamic neural network

This work presents a dynamic artificial neural network methodology, which classifies the proteins into their classes from their sequences alone: the lysosomal membrane protein classes and the various other membranes protein classes. In this paper, neural networks-based lysosomal-associated membrane protein type prediction system is proposed. Different protein sequence representations are fused to extract the features of a protein sequence, which includes seven feature sets; amino acid (AA) composition, sequence length, hydrophobic group, electronic group, sum of hydrophobicity, R-group, and dipeptide composition. To reduce the dimensionality of the large feature vector, we applied the principal component analysis. The probabilistic neural network, generalized regression neural network, and Elman regression neural network (RNN) are used as classifiers and compared with layer recurrent network (LRN), a dynamic network. The dynamic networks have memory, i.e. its output depends not only on the input but the previous outputs also. Thus, the accuracy of LRN classifier among all other artificial neural networks comes out to be the highest. The overall accuracy of jackknife cross-validation is 93.2% for the data-set. These predicted results suggest that the method can be effectively applied to discriminate lysosomal associated membrane proteins from other membrane proteins (Type-I, Outer membrane proteins, GPI-Anchored) and Globular proteins, and it also indicates that the protein sequence representation can better reflect the core feature of membrane proteins than the classical AA composition.

Variable structure dynamic artificial neural networks

We introduce a discrete-event artificial neural network structure inspired by biological neural networks. It includes dynamic components and has variable structure. The network’s topology and its dynamic components are modifiable and trainable for different applications. Such adaptation in the network’s parameters, structure, and dynamic components makes it easier to adapt to varying behaviors due to the problem’s structure than other types of networks. We demonstrate that this type of network structure can detect random changes in packet arrival rates in computer network traffic with possible applications in cyber security.

Twitter brand sentiment analysis: A hybrid system using n-gram analysis and dynamic artificial neural network

Twitter messages are increasingly used to determine consumer sentiment towards a brand. The existing literature on Twitter sentiment analysis uses various feature sets and methods, many of which are adapted from more traditional text classification problems. In this research, we introduce an approach to supervised feature reduction using n-grams and statistical analysis to develop a Twitter-specific lexicon for sentiment analysis. We augment this reduced Twitter-specific lexicon with brand-specific terms for brand-related tweets. We show that the reduced lexicon set, while significantly smaller (only 187 features), reduces modeling complexity, maintains a high degree of coverage over our Twitter corpus, and yields improved sentiment classification accuracy. To demonstrate the effectiveness of the devised Twitter-specific lexicon compared to a traditional sentiment lexicon, we develop comparable sentiment classification models using SVM. We show that the Twitter-specific lexicon is significantly more effective in terms of classification recall and accuracy metrics. We then develop sentiment classification models using the Twitter-specific lexicon and the DAN2 machine learning approach, which has demonstrated success in other text classification problems. We show that DAN2 produces more accurate sentiment classification results than SVM while using the same Twitter-specific lexicon.

Dynamic factor analysis and artificial neural network for estimating pan evaporation at multiple stations in northern Taiwan

Evaporation is an important reference for managers of water resources. This study proposes a hybrid model (BD) that combines back-propagation neural networks (BPNN) and dynamic factor analysis (DFA) to simultaneously precisely estimate pan evaporation at multiple meteorological stations in northern Taiwan through incorporating a large number of meteorological data sets into the estimation process. The DFA is first used to extract key meteorological factors that are highly related to pan evaporation and to establish the common trend of pan evaporation among meteorological stations. The BPNN is then trained to estimate pan evaporation with the inputs of the key meteorological factors and evaporation estimates given by the DFA. The BD model successfully inherits the advantages from the DFA and BPNN, and effectively enhances its generalization ability and estimation accuracy. The results demonstrate that the proposed BD model has good reliability and applicability in simultaneously estimating pan evaporation for multiple meteorological stations. Citation Chang, F.J., Sun, W., and Chung, C.H., 2013. Dynamic factor analysis and artificial neural network for estimating pan evaporation at multiple stations in northern Taiwan. Hydrological Sciences Journal, 58 (4), 813–825.

Observer-based output feedback control of discrete-time Luré systems with sector-bounded slope-restricted nonlinearities

SUMMARY Many well studied classes of dynamical systems such as actuator-constrained linear systems and dynamic artificial neural networks can be written as discrete-time Lure systems with sector-bounded and/or slope-restricted nonlinearities. Two types of observer-based output feedback control design methods are presented, compared, and analyzed with regard to robustness to model uncertainties and insensitivity to output disturbances. The controller designs are formulated in terms of LMIs that are solvable with standard software. The design equations are illustrated in numerical examples. Copyright © 2013 John Wiley & Sons, Ltd.

Using a dynamic artificial neural network for forecasting the volatility of a financial time series

The ability to obtain accurate volatility forecasts is an important issue for the financial analyst. In this paper, we use the DAN2 model, a multilayer perceptronand an ARCH model to predict the monthly conditional variance of stock prices.The results show that DAN2 model is more accurate for predicting in-sample andout-of-sample variance that the other considered models for the used data set. Thus, the value of this neural network as a predictive tool is demonstrated.

Artificial Neural Network Modeling of Ball Mill Grinding Process

Grinding consumes around 2% of the energy produced in the world but existing methods of milling are very inefficient and use only 5% of the input energy for real size reduction rest is consumed by machine itself. Chrome ores are comminute, filtered, pelletized and sintered to use into submerged arc furnace for ferrochrome production. Variation in ore properties affects the particle size distribution during milling. Artificial neural network based model is developed to predict the particle size distribution of ball mill product using grinding data available for difference in grindability of Sukinda chromite ores. Input variables for model were ball size, ball load, ball-ore ratio, grinding time. Output was particle size distribution (+75 μm, -75 μm, +38 μm; -38 μm). Three different kinds of mathematical models have been compared to predict the particle size distribution. Finally a neural network based model was found most accurate. Dynamic artificial neural network model does not require any material constant and optimizes the mathematical correlation with better accuracy in a dynamic process. This methodology can be used to develop an online system to predict the ball mill performance to improve the performance of grinding circuit in mineral, metal and cement industry.

MIMO-OFDM Channel Tracking Using a Dynamic ANN Topology

All the available algorithms for blind estimation namely constant modulus algorithm (CMA), Decision-Directed Algorithm (DDA/DFE) suffer from the problem of convergence to local minima. Also, if the channel drifts considerably, any DDA looses track of the channel. So, their usage is limited in varying channel conditions. The primary limitation in such cases is the requirement of certain overhead bits in the transmit framework which leads to wasteful use of the bandwidth. Also such arrangements fail to use channel state information (CSI) which is an important aid in improving the quality of reception. In this work, the main objective is to reduce the overhead imposed by the pilot symbols, which in effect reduces the system throughput. Also we formulate an arrangement based on certain dynamic Artificial Neural Network (ANN) topologies which not only contributes towards the lowering of the overhead but also facilitates the use of the CSI. A 2×2 Multiple Input Multiple Output (MIMO) system is simulated and the performance variation with different channel estimation schemes are evaluated. A new semi blind approach based on dynamic ANN is proposed for channel tracking in varying channel conditions and the performance is compared with perfectly known CSI and least square (LS) based estimation. Keywords—MIMO, Artificial Neural Network (ANN), CMA, LS, CSI.

Comparison of the ARMA, ARIMA, and the autoregressive artificial neural network models in forecasting the monthly inflow of Dez dam reservoir

Summary The goal of the present research is forecasting the inflow of Dez dam reservoir by using Auto Regressive Moving Average (ARMA) and Auto Regressive Integrated Moving Average (ARIMA) models while increasing the number of parameters in order to increase the forecast accuracy to four parameters and comparing them with the static and dynamic artificial neural networks. In this research, monthly discharges from 1960 to 2007 were used. The statistics related to first 42 years were used to train the models and the 5 past years were used to forecast. In ARMA and ARIMA models, the polynomial was derived respectively with four and six parameters to forecast the inflow. In the artificial neural network, the radial and sigmoid activity functions were used with several different neurons in the hidden layers. By comparing root mean square error (RMSE) and mean bias error (MBE), dynamic artificial neural network model with sigmoid activity function and 17 neurons in the hidden layer was chosen as the best model for forecasting inflow of the Dez dam reservoir. Inflow of the dam reservoir in the 12 past months shows that ARIMA model had a less error compared with the ARMA model. Static and Dynamic autoregressive artificial neural networks with activity sigmoid function can forecast the inflow to the dam reservoirs from the past 60 months.

Prediction and optimization of cure cycle of thick fiber-reinforced composite parts using dynamic artificial neural networks

Curing of thermoset-based composites experience substantial temperature overshoot, especially at the center of thick parts and large temperature gradient exists through the whole part due to large amount of heat released and low conductivity of the composite. This leads to non-uniformity of cure, residual stress and consequently composite cracks and possibly degradation of the polymer. The scope of this work is to optimize the cure cycle in order to improve the properties and gaining a relatively uniform part of composite, using trained recurrent artificial neural networks purposed for speeding up the repetitious model re-calls during the optimization process. Numerical results obtained based on the three-dimensional finite volume method is used to train the network. The optimization problem is aimed to develop multi-linear-stage cure cycles by minimizing the objective function that includes maximum temperature difference through the cure cycle with the constraints of maximum allowable temperature considering degradation temperature of the polymer, minimum temperature of cure initiation, and finally reaching a maximum final conversion. The results showed the effectiveness of the approach used in this study in terms of the optimization computation time.

Automated text classification using a dynamic artificial neural network model

Widespread digitization of information in today’s internet age has intensified the need for effective textual document classification algorithms. Most real life classification problems, including text classification, genetic classification, medical classification, and others, are complex in nature and are characterized by high dimensionality. Current solution strategies include Naïve Bayes (NB), Neural Network (NN), Linear Least Squares Fit (LLSF), k-Nearest-Neighbor (kNN), and Support Vector Machines (SVM); with SVMs showing better results in most cases. In this paper we introduce a new approach called dynamic architecture for artificial neural networks (DAN2) as an alternative for solving textual document classification problems. DAN2 is a scalable algorithm that does not require parameter settings or network architecture configuration. To show DAN2 as an effective and scalable alternative for text classification, we present comparative results for the Reuters-21578 benchmark dataset. Our results show DAN2 to perform very well against the current leading solutions (kNN and SVM) using established classification metrics.

Optimization of matrix tablets controlled drug release using Elman dynamic neural networks and decision trees

The main objective of the study was to develop artificial intelligence methods for optimization of drug release from matrix tablets regardless of the matrix type. Static and dynamic artificial neural networks of the same topology were developed to model dissolution profiles of different matrix tablets types (hydrophilic/lipid) using formulation composition, compression force used for tableting and tablets porosity and tensile strength as input data. Potential application of decision trees in discovering knowledge from experimental data was also investigated.Polyethylene oxide polymer and glyceryl palmitostearate were used as matrix forming materials for hydrophilic and lipid matrix tablets, respectively whereas selected model drugs were diclofenac sodium and caffeine. Matrix tablets were prepared by direct compression method and tested for in vitro dissolution profiles. Optimization of static and dynamic neural networks used for modeling of drug release was performed using Monte Carlo simulations or genetic algorithms optimizer. Decision trees were constructed following discretization of data.Calculated difference (f1) and similarity (f2) factors for predicted and experimentally obtained dissolution profiles of test matrix tablets formulations indicate that Elman dynamic neural networks as well as decision trees are capable of accurate predictions of both hydrophilic and lipid matrix tablets dissolution profiles. Elman neural networks were compared to most frequently used static network, Multi-layered perceptron, and superiority of Elman networks have been demonstrated. Developed methods allow simple, yet very precise way of drug release predictions for both hydrophilic and lipid matrix tablets having controlled drug release.

Depth of treatment sensitive noise resistant dynamic artificial neural networks model of recall in people with prosopagnosia

The Fusiform Face Area (FFA) is the brain region considered to be responsible for face recognition. Prosopagnosia is a brain disorder causing the inability to a recognise faces that is said to mainly affect the FFA. We put forward a model that simulates the capacity to retrieve label associated with faces and objects depending on the depth of treatment of the information. Akin to prosopagnosia, various localised "lesions" were inserted into the network in order to evaluate the degradation of performance. The network is first composed of a Feature Extracting Bidirectional Associative Memory (FEBAM-SOM) to represent the topological maps allowing the categorisation of all faces. The second component of the network is a Bidirectional Heteroassociative Memory (BHM) that links those representations to their semantic label. For the latter, specific semantic labels were used as well as more general ones. The inputs were images representing faces and various objects. Just like in the visual perceptual system, the images were pre-processed using a low-pass filter. Results showed that the network is able to associate the extracted map with the correct label information. The network is able to generalise and is robust to noise. Moreover, results showed that the recall performance of names associated with faces decrease with the size of lesion without affecting the performance of the objects. Finally, results obtained with the network are also consistent with human ones in that higher level, more general labels are more robust to lesion compared to low level, specific labels.

APPLYING k-SEPARABILITY TO COLLABORATIVE RECOMMENDER SYSTEMS

Most recommender systems have too many items to propose to too many users based on limited information. This problem is formally known as the sparsity of the ratings' matrix, because this is the structure that holds user preferences. This paper outlines a Collaborative Filtering Recommender System that tries to amend this situation. After applying Singular Value Decomposition to reduce the dimensionality of the data, our system makes use of a dynamic Artificial Neural Network architecture with boosted learning to predict user ratings. Furthermore we use the concept of k-separability to deal with the resulting noisy data, a methodology not yet tested in Recommender Systems. The combination of these techniques applied to the MovieLens datasets seems to yield promising results.

Urban Railway Bridge Fast Evaluation Based on Cloud Computing

Although modern bridge are carefully designed and well constructed, damage may occur in them due to unexpected causes. Currently, many different techniques have been proposed and investigated in bridge condition assessment. However, evaluation efficiency of condition assessment has not been paid much attention by the researchers. A fast evaluation of the urban railway bridge condition based on the cloud computing is presented. In this paper dynamic FE model and Artificial neural networks technique is applied to model updating. The cloud computing model provides the basis for fast analyses. It was found that when applied to the actually railway bridges, the proposed method provided results similar to those obtained by experts, but can improve efficiency of bridge

Predictions of SEP events by means of a linear filter and layer-recurrent neural network

Solar energetic particle (SEP) modelling has gained great interest in the community, specifically in connection with the safety of crews and the protection of technological systems of spacecraft situated outside the shielding of Earth's magnetosphere. Two models for the prediction of SEP events are presented in this paper. The models are based on a linear filter and on a special type of dynamic artificial neural network known as the layer-recurrent neural network. In this work they use as input the following parameters: the X-ray flare class for flares originating close to the centre of the solar disk; observed type II or IV radio bursts; and of the position angle, width, and linear speed of observed full or partial halo CMEs. The models are designed to provide forecasts of proton fluxes with energies exceeding 10 MeV at the L1 libration point.

Using artificial neural network models in stock market index prediction

Forecasting stock exchange rates is an important financial problem that is receiving increasing attention. During the last few years, a number of neural network models and hybrid models have been proposed for obtaining accurate prediction results, in an attempt to outperform the traditional linear and nonlinear approaches. This paper evaluates the effectiveness of neural network models which are known to be dynamic and effective in stock-market predictions. The models analysed are multi-layer perceptron (MLP), dynamic artificial neural network (DAN2) and the hybrid neural networks which use generalized autoregressive conditional heteroscedasticity (GARCH) to extract new input variables. The comparison for each model is done in two view points: Mean Square Error (MSE) and Mean Absolute Deviate (MAD) using real exchange daily rate values of NASDAQ Stock Exchange index.

A methodology for modeling batch reactors using generalized dynamic neural networks

This paper presents a methodology based on the application of dynamic artificial neural networks (DANNs) for modeling batch reactors. The network structure was designed by a specific method, called leave-one-out cross-validation. In order to reduce the number of input parameters, the multiway principal component analysis (MPCA) was employed. As a case study, sequencing batch reactor was selected to examine the suggested procedure. The results of DANN model were compared to the experimental data, extracted from the literature. Different statistical tools were used as the evaluation criteria for this comparison. The relative error of training and testing sets were 2.11% and 2.6%, respectively. The regression between the network outputs and the experimental data was more than 0.95. Therefore, the model developed in this work has an acceptable generalization capability and accuracy. In addition, it was proved that the implementation of MPCA with dynamic neural network could enhance the model performance. Furthermore, the comparison between the DANN model predictions with those of a mechanistic model revealed that the recommended model was over two and half times more accurate.

Non-linear wave data assimilation with an ANN-type wind-wave model and Ensemble Kalman Filter (EnKF)

Non-linear data assimilation for a wind-wave dynamical surrogate model in a reduced space is presented. A dynamic artificial neural network is used for surrogate modeling. It provides a fast emulation of a wind-wave model which is used for the evaluation of the system states during a small period of time. The system state consists of wave height and wave direction in the reduced space which is affected by the reduced space wind field. The projection from the full space to the reduced one is performed by a principal component analysis. Ensemble methods require the evaluation of dynamics for a large number of statistical ensembles, so coupling this surrogate (instead of a full model) with an Ensemble Kalman Filter (EnKF) leads to computational efficiency. Application of the procedure is demonstrated through 6 month hindcast study of wind waves over the Caspian Sea using the third-generation wave model and the analysis of the ECMWF wind field. The trained network is embedded into the stochastic environment. Then, the EnKF is used to find estimate of the system states. Experiments show that the proposed data assimilation technique can correct the prediction of the wind-waves requiring just a modest execution time.

Measuring effectiveness of a dynamic artificial neural network algorithm for classification problems

Classification is the process of assigning an object to one of a set of classes based on its attributes. Classification problems have been examined in fields as diverse as biology, medicine, business, image recognition, and forensics. Developing more accurate and widely applicable classification methods has significant implications in these and many other fields. This paper presents a dynamic artificial neural network (DAN2) as an alternate approach for solving classification problems. We show DAN2 to be an effective approach and compare its performance with linear discriminant analysis, quadratic discriminant analysis, k-nearest neighbor algorithms, support vector machines, and traditional artificial neural networks using benchmark and real-world application data sets. These data sets vary in the number of classes (two vs. multiple) and the source of the data (synthetic vs. real-world). We found DAN2 to be a very effective classification method for two-class data sets with accuracy improvements as high as 37.2% when compared to the other methods. We also introduce a hierarchical DAN2 model for multiple class data sets that shows marked improvements (up to 89%) over all other methods, and offers better accuracy in all cases.