Nonlinear Neural Network Model Research Articles

Computer-Aided Design of Antimicrobial Peptides

An increasing number of reported cases of drug resistant Staphylococcus aureus and Pseudomonas aeruginosa, demonstrate the urgent need for new therapeutics that are effective against such and other multi-drug resistant bacteria. Antimicrobial peptides have for two decades now been looked upon as interesting leads for development of new therapeutics combating these drug resistant microbes. High-throughput screening of peptide libraries have generated large amounts of information on peptide activities. However, scientists still struggle with explaining the specific peptide motifs resulting in antimicrobial activity. Consequently, the majority of peptides put into clinical trials have failed at some point, underlining the importance of a thorough peptide optimization. An important tool in peptide design and optimization is quantitative structure-activity relationship (QSAR) analysis, correlating chemical parameters with biological activities of the peptide, using statistical methods. In this review we will discuss two different in silico strategies of computer-aided antibacterial peptide design, a linear correlation model build as an extension of traditional principal component analysis (PCA) and a non-linear artificial neural network model. Studies on structurally diverse peptides, have concluded that the PCA derived model are able to guide the antibacterial peptide design in a meaningful way, however requiring rather a high homology between the peptides in the test-set and the in silico library, to ensure a successful prediction. In contrast, the neural network model, though significantly less explored in relation to antimicrobial peptide design, has proven extremely promising, demonstrating impressive prediction success and ranking of random peptide libraries correlating well with measured activities. Keywords: P. aeruginosa, Antimicrobial Peptides, Quantitative Structure-Activity Relationships, Prediction of activity, Partial Least Square Projections to Latent Structures, Artificial Neural Network

Exploration of lead factors affecting service recovery

This research attempts to explore the influence of empowerment on service recovery from the viewpoint of managers and also to probe into the possible differences in the adoption of service recovery strategies in different corporate cultures. Linear multivariate data analysis and nonlinear fuzzy neural network are combined to analyze data and verify the proposed hypotheses. Through the investigation into Chinese and Western food chain stores, it is shown that the more empowered employees will adopt more active failure recovery strategies; and tougher corporate cultures tend to have passive service recovery strategies, whereas minor corporate cultures tend to have active service recovery strategies. Customer relationship involvement reveals positive influences on the adoption of recovery strategies. The characteristic of this research is that, on the one hand, through the empirical conclusion of the nonlinear fuzzy neural network model, we not only measure the relationship among the variables more precisely, but also have less restrictive conditions. Also, according to organizational management factors, this research proposes and examines the influencing factors affecting service recovery strategies. It explores the responses to service recovery from the viewpoint of internal prevention in the organization, which is different from past research that focussed mostly upon consumers' views.

Service recovery expectation model – from the perspectives of consumers

This paper explores the influence of the major attributes of consumers, such as personality traits, gender, number of consumptions, and levels of involvement, on service recovery expectations. A combination of linear multivariate statistical analysis and nonlinear fuzzy neural network models is further used to analyse data and validate hypotheses. Consumers of middle-to-high-end Chinese restaurant chains are sampled as the interactions between these restaurants and consumers are frequent. The empirical research finds that old customers have lower expectations of service recovery than new ones and male consumers have lower expectations than female consumers. Extrovert-type consumers have lower expectations of service recovery than introvert-type consumers. There is a positive correlation between consumers' involvement and service recovery expectation. As this paper examines the expectations of service recovery from the perspectives of consumers, it aims to help companies use pre-processing of service recovery as a precautionary measure, rather than as a remedial measure as mostly seen. Different from the vast majority of literature that deals with the issue by focusing on western enterprises, this paper uses small-and-medium enterprises in Taiwan and constructs management implications accordingly.

Service failure and consumer switching behaviors: Evidence from the insurance industry

This paper analyzes consumer switching behaviors from the perspectives of consumer psychology and interactions with the society, such as consumer’s emotional intelligence, relationship involvement and personality traits. We survey life insurance policyholders in Taiwan and Hong Kong. Our results show that when service failures are more serious, consumers with higher emotional intelligence exhibit lower intention to switch than those with lower emotional intelligence. Similarly, the more serious the service failures, the more the consumers with internal locus-of-control orientation exhibit intention to change their existing consumption than those with external locus-of-control orientation. Further for more serious the service failures, the higher the relationship involvement with customers, the more their intention to switch is reduced. The unique contribution of this paper is that it combines multivariate statistical analysis and a non-linear neural fuzzy network model structure to verify the collected data. The paper concludes with a discussion of management implications and recommendations for future studies.

QSPR study for the prediction of half-wave potentials of benzoxazines by heuristic method and radial basis function neural network

Abstract The half-wave potential (E1/2) is an important electrochemical property of organic compounds. In this work, a quantitative structure-property relationship (QSPR) analysis has been conducted on the half-wave reduction potential (E1/2) of 40 substituted benzoxazines by means of both a heuristic method (HM) and a non-linear radial basis function neural network (RBFNN) modeling method. The statistical parameters provided by the HM model (R2 =0.946; F=152.576; RMSCV=0.0141) and the RBFNN model (R2=0.982; F=1034.171 and RMS =0.0209) indicated satisfactory stability and predictive ability. The obtained models showed that benzoxazines with larger Min valency of a S atom (MVSA), lower Relative number of H atom (RNHA) and Min n-n repulsion for a C-H bond (MnnRCHB) and Minimal Electrophilic Reactivity Index for a C atom (MERICA) can be more easily reduced. This QSPR approach can contribute to a better understanding of structural factors of the organic compounds that contribute to the E1/2, and can be useful in predicting the E1/2 of other compounds.

Service recovery model: the integrated view

Through the integrated analytical view of overall cultural values and individual emotional intelligence (EI), psychological empowerment and management involvement, this research explores influences on service recovery. It combines the methods of linear multivariate data analysis and nonlinear fuzzy neural network methods to analyze data and verify the hypotheses. Through an investigation on the subsidiaries of financial groups in Taiwan and Hong Kong, the empirical results showed that the employees with lower EI would be affected by the involvement of management and adopt different service recovery measures; for the employees having traditional eastern cultural values, because of the involvement of management, they would adopt different service recovery measures; when the employees' perceived psychological empowerment level was higher, they were more likely to manage active recovery measures. The characteristic of this research combines the empirical methods of linear multivariate data analysis and the nonlinear fuzzy neural network model. There were rare related studies in the past and in addition, in the comparative studies of the related issues in the past, there were rare focuses on enterprises with different nationalities but sharing the same specific language, which was worthy of further exploration.

Multi step ahead prediction of north and south hemisphere sun spots chaotic time series using focused time lagged recurrent neural network model

Multi-Step ahead prediction of a chaotic time series is a difficult task that has attracted increasing interest in recent years. The interest in this work is the development of nonlinear neural network models for the purpose of building multistep ahead prediction of North and South hemisphere sunspots chaotic time series. In the literature there is a wide range of different approaches but their success depends on the predicting performance of the individual methods. Also the most popular neural models are based on the statistical and traditional feed forward neural networks. But it is seen that this kind of neural model may present some disadvantages when long-term prediction is required. In this paper focused time lagged recurrent neural network (FTLRNN) model with gamma memory is developed not only for short-term but also for long-term prediction which allows to obtain better predictions of northern and southern chaotic time series in future. The authors experimented the performance of this FTLRNN model on predicting the dynamic behavior of typical northern and southern sunspots chaotic time series. Static MLP model is also attempted and compared against the proposed model on the performance measures like mean squared error (MSE), Normalized mean squared error (NMSE) and Correlation Coefficient (r). The standard back propagation algorithm with momentum term has been used for both the models. The various parameters like number of hidden layers, number of processing elements in the hidden layer, step size, the different learning rules, the various transfer functions like tanh, sigmoid, linear-tanh and linear sigmoid, different error norms L1, L2 (Euclidean), L3, L4, L5 and L∞, and different combination of training and testing samples are exhaustively varied and experimented for obtaining the optimal values of performance measures. The obtained results indicates the superior performance of estimated dynamic FTLRNN based model with gamma memory over the static MLP NN in various performance metrics. In addition, the output of proposed FTLRNN neural network model with gamma memory closely follows the desired output for multi- step ahead prediction for all the chaotic time series considered in the study.

Novel FTLRNN with Gamma Memory for Short-Term and Long-Term Predictions of Chaotic Time Series

Multistep ahead prediction of a chaotic time series is a difficult task that has attracted increasing interest in the recent years. The interest in this work is the development of nonlinear neural network models for the purpose of building multistep chaotic time series prediction. In the literature there is a wide range of different approaches but their success depends on the predicting performance of the individual methods. Also the most popular neural models are based on the statistical and traditional feed forward neural networks. But it is seen that this kind of neural model may present some disadvantages when long-term prediction is required. In this paper focused time-lagged recurrent neural network (FTLRNN) model with gamma memory is developed for different prediction horizons. It is observed that this predictor performs remarkably well for short-term predictions as well as medium-term predictions. For coupled partial differential equations generated chaotic time series such as Mackey Glass and Duffing, FTLRNN-based predictor performs consistently well for different depths of predictions ranging from short term to long term, with only slight deterioration after k is increased beyond 50. For real-world highly complex and nonstationary time series like Sunspots and Laser, though the proposed predictor does perform reasonably for short term and medium-term predictions, its prediction ability drops for long term ahead prediction. However, still this is the best possible prediction results considering the facts that these are nonstationary time series. As a matter of fact, no other NN configuration can match the performance of FTLRNN model. The authors experimented the performance of this FTLRNN model on predicting the dynamic behavior of typical Chaotic Mackey-Glass time series, Duffing time series, and two real-time chaotic time series such as monthly sunspots and laser. Static multi layer perceptron (MLP) model is also attempted and compared against the proposed model on the performance measures like mean squared error (MSE), Normalized mean squared error (NMSE), and Correlation Coefficient (r). The standard back-propagation algorithm with momentum term has been used for both the models.

Antecedents of employee involvement with the comparative model

Instead of discussing employee’s performance that previous researchers emphasized in their studies on the work behavior of employees, this study analyzes employee involvement from the viewpoint of their emotional labor, organizational culture, and the intimacy relationships between employees and their supervisors. Large-scale financial holding conglomerates in Taiwan and subsidiaries of Hong Kong-based conglomerates were selected as the populations for the research. The result of the empirical analysis indicates that emotional labor has a significantly positive effect on the involvement of employees; the stronger the intimacy relationship between employees and their supervisors is, the higher the employee involvement will be, strong culture has a significantly negative effect on the involvement of employees, and there is a significant difference in a number of dimensions among some Chinese enterprises in Taiwan & Hong Kong. This study has the following characteristics in comparison with the previous studies: (1) The theoretical structure of this study is formed based on an integrated prospect of internal organizational structure, personal factors of employees and their interpersonal relationship; (2) the study focuses on a comparative study of regional enterprises, which was seldom emphasized in literature; (3) the study uses the Non-Linear Fuzzy Neural Network Model and multivariate analysis approach as tools for the research and may contribute more to the theory in this field.

Computational chemistry study of 3D‐structure‐function relationships for enzymes based on Markov models for protein electrostatic, HINT, and van der Waals potentials

In a significant work, Dobson and Doig (J Mol Biol 2003, 330, 771) illustrated protein prediction as enzymatic or not from spatial structure without resorting to alignments. They used 52 protein features and a nonlinear support vector machine model to classify more than 1000 proteins collected from the PDB with a 77% overall accuracy. The most useful features were: the secondary-structure content, the amino acid frequencies, the number of disulphide bonds, and the largest cleft size. Working on the same dataset used by D&D, in this article we reported a good and simple model, based on the Markov chain models (MCM), to classify protein 3D structures as enzymatic or not, taking into consideration the spatial structure without resorting to alignments. Here we define, for the first time, a general MCM to calculate the electrostatic potential, molecular vibrations, van der Waals (vdw) interactions, and hydrophobic interactions (HINT) and use them in comparative studies of potential fields and/or protein function prediction. The dataset is composed of 1371 proteins divided into 689 enzymes and 682 nonenzymes, all proteins were collected from the PDB. The best model we found was a linear model carried out with the linear discriminant analysis; it was able to classify 74.18% of the proteins using only two electrostatic potentials. In the work described here, we define 3D-HINT potentials (mu(k)) and use them for the first time to derive a classifier for protein enzymes. We analyzed ROC curves, domain of applicability, parametric assumptions, desirability maps, and also tested other nonlinear artificial neural network models which did not improve the linear model. In closing, this MCM allows a fast calculation and comparison of different potentials deriving into accurate protein 3D structure-function relationships, notably simpler than the previous.

Predictive modeling for wastewater applications: Linear and nonlinear approaches

This study compares the multi-period predictive ability of linear ARIMA models to nonlinear time delay neural network models in water quality applications. Comparisons are made for a variety of artificially generated nonlinear ARIMA data sets that simulate the characteristics of wastewater process variables and watershed variables, as well as two real-world wastewater data sets. While the time delay neural network model was more accurate for the two real-world wastewater data sets, the neural networks were not always more accurate than linear ARIMA for the artificial nonlinear data sets. In some cases of the artificial nonlinear data, where multi-period predictions are made, the linear ARIMA model provides a more accurate result than the time delay neural network. This study suggests that researchers and practitioners should carefully consider the nature and intended use of water quality data if choosing between neural networks and other statistical methods for wastewater process control or watershed environmental quality management.

Signal restoration from atmospheric degradation in terahertz spectroscopy

This paper presents a method of restoring signals in terahertz spectroscopy by removing the distortion from the observed terahertz signals. The distortion is generated by the absorption and scattering of gas molecules and water vapor in the atmosphere during the transmission of terahertz beams through the air from the source to the spectrometer. Such atmospheric degradation causes spurious spectral dips and peaks in the terahertz spectrum, which often obscure the spectral peaks specific to the material of interest. This fact makes it challenging to measure the terahertz spectroscopic signatures of objects in a humid air environment, even at a short distance of approximately 1 m. A terahertz signal restoration filter based on a nonlinear artificial neural network model effectively removes noisy absorption peaks in terahertz spectra caused by atmospheric degradation.

Simulation of torsional shear test results with neuro-fuzzy control system

In the first part of this study, a series of stress-controlled hollow cylinder cyclic torsional triaxial shear tests were conducted on loose to medium dense saturated samples of clean Toyoura sand to investigate its liquefaction behavior. A uniform cyclic sinusoidal loading at a 0.1 Hz frequency was applied to air-pluviated samples where confining pressure and relative density was varied. Cyclic shear stress–strain changes, the number of cycles to reach liquefaction and pore pressure variations were recorded. Results indicate that the liquefaction resistances of uniform sands are significantly affected by the method of sample preparation and initial conditions. In the second phase, an adaptive neuro-fuzzy inference methodology was applied for the simulation of nonlinear mapping among radial stress conditions, relative density, and number of load repetitions necessary for the liquefaction. The neuro-fuzzy model included a fuzzy approximate reasoning through a Sugeno fuzzy inference system (FIS) and the input space was fuzzified by a grid-partitioning technique. A hybrid-learning algorithm was selected for the adaptation of model parameters. Furthermore, linear–nonlinear regression analyses and neural network models were employed to observe relative performances. It is concluded based on the findings that the neuro-fuzzy control system exhibited a superior performance as compared to other employed methods.

Drought Forecasting Using a Hybrid Stochastic and Neural Network Model

Treating the occurrence and severity of droughts as random, a hybrid model, combining a linear stochastic model and a nonlinear artificial neural network (ANN) model, is developed for drought forecasting. The hybrid model combines the advantages of both stochastic and ANN models. Using the Standardized Precipitation Index series, the hybrid model as well as the individual stochastic and ANN models were applied to forecast droughts in the Kansabati River basin in India, and their performances were compared. The hybrid model was found to forecast droughts with greater accuracy.

A nonlinear dynamic artificial neural network model of memory

Nonlinearity and dynamics in psychology are found in various domains such as neuroscience, cognitive science, human development, etc. However, the models that have been proposed are mostly of a computational nature and ignore dynamics. In those models that do include dynamic properties, only fixed points are used to store and retrieve information, leaving many principles of nonlinear dynamic systems (NDS) aside; for instance, chaos is often perceived as a nuisance. This paper considers a nonlinear dynamic artificial neural network (NDANN) that implements NDS principles while also complying with general neuroscience constraints. After a theoretical presentation, simulation results will show that the model can exhibit multi-valued, fixed-point, region-constrained attractors and aperiodic (including chaotic) behaviors. Because the capabilities of NDANN include the modeling of spatiotemporal chaotic activities, it may be an efficient tool to help bridge the gap between biological memory neural models and behavioral memory models.

Improved and More-Rapid History Matching With a Nonlinear Proxy and Global Optimization

This article, written by Technology Editor Dennis Denney, contains highlights of paper SPE 101933, "Improved and More- Rapid History Matching With a Nonlinear Proxy and Global Optimization," by A.S. Cullick, SPE, W.D. Johnson, SPE, and G. Shi, Landmark Graphics, prepared for the 2006 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 24–27 September. Generating reservoir-simulation models that match field production data has been a long-time industry challenge. Two workflows are presented for assisting history match. One workflow minimizes the misfit between simulated and historical data with a global optimizer. Another workflow trains a comprehensive nonlinear-proxy model with a small set of numerical simulations from experimental design to reduce the number of numerical simulations. Introduction History matching is, by its nature, an ill-posed optimization problem with many unknown reservoir parameters that could be adjusted to achieve a match against a relatively small amount of measured data at wells. The most common method of history match is to execute many simulations one at a time, changing a few parameters in a trial-and-error fashion. Often, a reservoir-simulation history match might require months of effort and many simulations to achieve a single model that neglects model nonuniqueness and often may not be a good predictor of future field performance. Many automated history-match algorithms have been investigated. Early assisted history matching emphasized deterministic gradient-optimization algorithms that require complete derivatives of the production response with respect to reservoir parameters. Subsequently, the emphasis evolved to faster methods to compute the sensitivity coefficients, such as streamline simulation, with emphasis on preserving uncertainty through the integration of geostatistics. Recent developments with adjoint models have renewed interest in sensitivity-based algorithms. However, none of these methods have been taken up widely by practitioners. With black-box optimization, the number of required simulations to achieve a match can be very large, numbered in the hundreds of simulations. If a single simulation model takes many hours to execute, there is good incentive to reduce the number of simulations for the optimization, even with the use of a distributed-computing system. This study used a global optimizer, while seeking to reduce the number of simulations with a fast proxy model. The use of proxy-regression models has been suggested to reduce the number of required simulations, but analytical-response models may not be sufficiently robust to represent the nonlinear reservoir production and pressure vs. time profile data. The use of neural networks as nonlinear-proxy models also has been suggested. In this paper, a history-match workflow is presented that has a robust, nonlinear neural-network proxy model to improve a history match and reduce the number of finite-difference simulations required.

The effect of knowledge sharing model

Three elements - organizational structure characteristics, organizational culture, and interunit interaction - that affect knowledge sharing are raised in this study based on the structure of organization theories and the interaction among units of an organization. By carrying out empirical research on five hi-tech industries in Taiwan and verifying hypotheses with the non-linear fuzzy neural network model, this study finds that the lower the formalization of an organizational structure is, the greater the knowledge sharing among units of an organization will be, while the higher the complexity of an organizational structure is, the lower the knowledge sharing among units of an organization will be. Trust and commitment among units are important for facilitating knowledge sharing among units, and creative and supporting characteristics of organizational culture are beneficial for the implementation of knowledge sharing activities.

Evolutionary optimization of classifiers and features for single-trial EEG Discrimination

BackgroundState-of-the-art signal processing methods are known to detect information in single-trial event-related EEG data, a crucial aspect in development of real-time applications such as brain computer interfaces. This paper investigates one such novel approach, evaluating how individual classifier and feature subset tailoring affects classification of single-trial EEG finger movements. The discrete wavelet transform was used to extract signal features that were classified using linear regression and non-linear neural network models, which were trained and architecturally optimized with evolutionary algorithms. The input feature subsets were also allowed to evolve, thus performing feature selection in a wrapper fashion. Filter approaches were implemented as well by limiting the degree of optimization.ResultsUsing only 10 features and 100 patterns, the non-linear wrapper approach achieved the highest validation classification accuracy (subject mean 75%), closely followed by the linear wrapper method (73.5%). The optimal features differed much between subjects, yet some physiologically plausible patterns were observed.ConclusionHigh degrees of classifier parameter, structure and feature subset tailoring on individual levels substantially increase single-trial EEG classification rates, an important consideration in areas where highly accurate detection rates are essential. Also, the presented method provides insight into the spatial characteristics of finger movement EEG patterns.

Combined Hydraulic and Black-Box Models for Flood Forecasting in Urban Drainage Systems

Rapid urbanization and its implications for both water quality issues and floods have increased the need for modeling of urban drainage systems. Many operational models are based on deterministic solutions of hydraulic equations. Improving such models by adding a “black-box” component to deal with any systematic structure in the residuals is proposed. In this study, a conventional deterministic stormwater drainage network model is first developed for a rapidly developing catchment using the HYDROWORKS (now called Infoworks) package, from Wallingford Software in the United Kingdom. However, despite the generally satisfactory results, the HYDROWORKS model tended to underestimate the flow volume. In this paper, a black-box or “systems” model is fitted to the hydraulic urban drainage model in order to improve its overall efficiency. A study was conducted of suitable black-box models, which included the nonlinear artificial neural network model (ANN), and the linear time series models of Box and Jenkins in 197...

Comparison of Gompertz and nEural Network Models of Broiler Growth

Neural networks offer an alternative to regression analysis for biological growth modeling. Very little research has been conducted to model animal growth using artificial neural networks. Twenty-five male chicks (Ross × Ross 308) were raised in an environmental chamber. Body weights were determined daily and feed and water were provided ad libitum. The birds were fed a starter diet (23% CP and 3,200 kcal of ME/kg) from 0 to 21 d, and a grower diet (20% CP and 3,200 kcal of ME/ kg) from 22 to 70 d. Dead and female birds were not included in the study. Average BW of 18 birds were used as the data points for the growth curve to be modeled. Training data consisted of alternate-day weights starting with the first day. Validation data consisted of BW at all other age periods. Comparison was made between the modeling by the Gompertz nonlinear regression equation and neural network modeling. Neural network models were developed with the Neuroshell Predictor. Accuracy of the models was determined by mean square error (MSE), mean absolute deviation (MAD), mean absolute percentage error (MAPE), and bias. The Gompertz equation was fit for the data. Forecasting error measurements were based on the difference between the model and the observed values. For the training data, the lowest MSE, MAD, MAPE, and bias were noted for the neural-developed neural network. For the validation data, the lowest MSE and MAD were noted with the genetic algorithm-developed neural network. Lowest bias was for the neural-developed network. As measured by bias, the Gompertz equation underestimated the values whereas the neural- and genetic-developed neural networks produced little or no overestimation of the observed BW responses. Past studies have attempted to interpret the biological significance of the estimates of the parameters of an equation. However, it may be more practical to ignore the relevance of parameter estimates and focus on the ability to predict responses.