Multilayer Perceptron

This study is concerned with understanding of the formation of ore deposits (precious and base metals) and contributes to the exploration and discovery of new occurrences using artificial neural networks. From the different digital data sets available in BRGM's GIS ...

This study is concerned with understanding of the formation of ore deposits (precious and base metals) and contributes to the exploration and discovery of new occurrences using artificial neural networks. From the different digital data sets available in BRGM's GIS ...

This paper surveys in a tutorial fashion the recent history of universal learning machines starting with the multilayer perceptron.The big push in recent years has been on the design of universal learning machines using optimization methods linear in the parameters, such as the Echo State Network, the Extreme Learning Machine and the Kernel Adaptive filter. We call this class of learning machines convex universal learning machines or CULMs. The purpose of the paper is to compare the methods behind these CULMs, highlighting their features using concepts of vector spaces (i.e. basis functions and projections), which are easy to understand by the computational intelligence community. We illustrate how two of the CULMs behave in a simple example, and we conclude that indeed it is practical to create universal mappers with convex adaptation, which is an improvement over backpropagation.

In this paper, we investigate the effectiveness of a financial time-series forecasting strategy which exploits the mul- tiresolution property of the wavelet transform. A financial series is decomposed into an over complete, shift invariant scale-related representation. In transform space, each individual wavelet series is modeled by a separate multilayer perceptron (MLP). To better utilize the detailed information in the lower scales of wavelet coef- ficients (high frequencies) and general (trend) information in the higher scales of wavelet coefficients (low frequencies), we applied the Bayesian method of automatic relevance determination (ARD) to choose short past windows (short-term history) for the inputs to the MLPs at lower scales and long past windows (long-term history) at higher scales. To form the overall forecast, the indi- vidual forecasts are then recombined by the linear reconstruction property of the inverse transform with the chosen autocorrelation shell representatio...

This work presents system identification using neural network approaches for modelling a laboratory based twin rotor multi-input multi-output system (TRMS). Here we focus on a memetic algorithm based approach for training the multilayer perceptron neural network (NN) applied to nonlinear system identification. In the proposed system identification scheme, we have exploited three global search methods namely genetic algorithm (GA), Particle Swarm Optimization (PSO) and differential evolution (DE) which have been hybridized with the gradient descent method i.e. the back propagation (BP) algorithm to overcome the slow convergence of the evolving neural networks (EANN). The local search BP algorithm is used as an operator for GA, PSO and DE. These algorithms have been tested on a laboratory based TRMS for nonlinear system identification to prove their efficacy.

This work addresses the real time control of the Khepera mobile robot [1] navigation in a maze with reflector walls. Boolean Neural Networks such as RAM [2] and GSN [3] models are applied to drive the vehicle, following a light source, while avoiding obstacles. Both neural networks are implemented with simple logic and arithmetic functions (NOT, AND, OR, Addition, and Comparison), aiming to improve system speed. The results obtained are compared with two other control strategies: Multi-layer Perceptron (MLP) [4] and Fuzzy Logic [5].

Adsorption of Fe(III) from water was optimized using response surface methodology and artificial neural network. The initial concentration of Fe(III), contact time, and concentration of adsorbent were defined as input variables, while the percentage of adsorbed Fe(III) was labeled as output variable. Bentonite clay served as an adsorbent. The correlation coefficient, root mean square error, and mean absolute error were calculated after fitting the experimental data by the second-order polynomial model at the central composite design (CCD). The multilayer perceptron with architecture of 3-9-1 provided the best performance. A logistic sigmoid was applied as an activation function in the hidden layer, while in the output layer a linear function was used.

In this paper we describe a training method for one hidden layer multilayer perceptron classifier which is based on the idea of support vector machines (SVM's). An upper bound on the Vapnik-Chervonenkis (VC) dimension is iteratively minimized over the interconnection matrix of the hidden layer and its bias vector. The output weights are determined according to the support vector method, but without making use of the classifier form which is related to Mercer's condition. The method is illustrated on a two-spiral classification problem.

The ability to forecast the future based on past data is a key tool to support individual and organizational decision making. In particular, the goal of Time Series Forecasting (TSF) is to predict the behavior of complex systems by looking only at past patterns of the same phenomenon. In recent years, several works in the literature have adopted Evolutionary Artificial Neural Networks (EANN) for TSF. In this work, we propose a novel EANN approach, where a weighted n-fold validation fitness scheme is used to build an ensemble of neural networks, under four different combination methods: mean, median, softmax and rank-based. Several experiments were held, using six real-world time series with different characteristics and from distinct domains. Overall, the proposed approach achieved competitive results when compared with a non-weighted n-fold EANN ensemble, the simpler 0-fold EANN and also the popular Holt-Winters statistical method.

This paper presents a recently introduced Kernel Machine, called Geometrical Kernel Machine, used to predict disruptive events in nuclear fusion reactors. The algorithm proposed to construct the Kernel Machine is able to automatically determine both the number of neurons and the synaptic weights of a Multilayer Perceptron neural network with a single hidden layer. It has been demonstrated that the resulting network is able to classify any finite set of patterns defined in a real domain. The prediction problem has been here modeled as a two classes classification problem. The geometrical interpretation of the network equations allows us both to develop the disruption predictor and to manage the so called ageing of the kernel machine. In fact, using the same kernel machine, a novelty detection system has been integrated in the predictor, increasing the overall system performance, and the reliability of the predictor.

This paper introduces and evaluates the use of Gaussian mixture models (GMMs) for multiple limb motion classification using continuous myoelectric signals. The focus of this work is to optimize the configuration of this classification scheme. To that end, a complete experimental evaluation of this system is conducted on a 12 subject database. The experiments examine the GMMs algorithmic issues including the model order selection and variance limiting, the segmentation of the data, and various feature sets including time-domain features and autoregressive features. The benefits of postprocessing the results using a majority vote rule are demonstrated. The performance of the GMM is compared to three commonly used classifiers: a linear discriminant analysis, a linear perceptron network, and a multilayer perceptron neural network. The GMM-based limb motion classification system demonstrates exceptional classification accuracy and results in a robust method of motion classification with low computational load.

Feature extraction has been always mutually studied for exploratory data projection and for classification. Feature extraction for exploratory data projection aims for data visualization by a projection of a high-dimensional space onto two or three-dimensional space, while feature extraction for classification generally requires more than two or three features. Therefore, feature extraction paradigms for exploratory data projection are not commonly employed for classification and vice versa. We study extraction of more than three features, using neural network (NN) implementation of Sammon's nonlinear mapping to be applied for classification.

In corporate data mining applications, cost-sensitive learning is firmly established for predictive classification algorithms. Conversely, data mining methods for regression and time series analysis generally disregard economic utility and apply simple accuracy measures. Methods from statistics and computational intelligence alike minimise a symmetric statistical error, such as the sum of squared errors, to model ordinary least squares predictors. However, applications in business elucidate that real forecasting problems contain non-symmetric errors. The costs arising from over-versus underprediction are dissimilar for errors of identical magnitude, requiring an ex-post correction of the prediction to derive valid decisions. To reflect this, an asymmetric cost function is developed and employed as the objective function for neural network training, deriving superior forecasts and a cost efficient decision. Experimental results for a business scenario of inventory-levels are computed using a multilayer perceptron trained with different objective functions, evaluating the performance in competition to statistical forecasting methods.

In soccer, scoring goals is a fundamental objective which depends on many conditions and constraints. Considering the RoboCup soccer 2D-simulator, this paper presents a data mining-based decision system to identify the best time and direction to kick the ball towards the goal to maximize the overall chances of scoring during a simulated soccer match. Following the CRISP-DM methodology, data for modeling were extracted from matches of major international tournaments (10691 kicks), knowledge about soccer was embedded via transformation of variables and a Multilayer Perceptron was used to estimate the scoring chance. Experimental performance assessment to compare this approach against previous LDA-based approach was conducted from 100 matches. Several statistical metrics were used to analyze the performance of the system and the results showed an increase of 7.7% in the number of kicks, producing an overall increase of 78% in the number of goals scored.

The latency of current mobile devices' touchscreens is around 100ms and has widely been explored. Latency down to 2ms is noticeable, and latency as low as 25ms reduces users' performance. Previous work reduced touch latency by extrapolating a finger's movement using an ensemble of shallow neural networks and showed that predicting 33ms into the future increases users' performance. Unfortunately, this prediction has a high error. Predicting beyond 33ms did not increase participants' performance, and the error affected the subjective assessment. We use more recent machine learning techniques to reduce the prediction error. We train LSTM networks and multilayer perceptrons using a large data set and regularization. We show that linear extrapolation causes an 116.7% higher error and the previously proposed ensembles of shallow networks cause a 26.7% higher error compared to the LSTM networks. The trained models, the data used for testing, and the source code is available on GitHub.

Telemarketing is a promotion that is considered effective for promoting a product to consumers by telephone, other than that telemarketing is easier to accept because of its direct nature of offering products to consumers. Telemarketing is also considered to help increase a company's revenue. The problem of predicting the success of a bank's telemarketing data must be done using machine learning techniques. Machine learning used in the available historical data is a bank dataset of 45211 instances at 17 features using the multilayer perceptron algorithm (MLP) with resampling. The use of resampling aims to balance the unbalanced data resulting in an accuracy value of 90.18% and a ROC of 0.89%. Meanwhile, if the data resampling is not used in the multilayer perceptron (MLP) algorithm, the accuracy value is 88.6 and ROC is 0.88%. The use of resampling data becomes more effective and results in higher accuracy values.

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 longterm 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 L 1 ,L 2 (Euclidean), L 3, L 4 ,L 5 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.

Recent work on extracting features of gaps in handwritten text allows a classification into inter-word and intraword classes using suitable classification techniques. In this paper, we apply 5 different supervised classification algorithms from the machine learning field on both the original dataset and a dataset with the best features selected using mutual information. The classifiers are compared by employing McNemar's test. We find that SVMs and MLPs outperform the other classifiers and that preprocessing to select features works well.

The correct diagnosis of breast cancer is one of the major problems in the medical field. From the literature it has been found that different pattern recognition techniques can help them to improve in this domain. These techniques can help doctors form a second opinion and make a better diagnosis. In this paper we present a novel improvement in neural network training for pattern classification. The proposed training algorithm is inspired by the biological metaplasticity property of neurons and Shannon's information theory. During the training phase the Artificial metaplasticity Multilayer Perceptron (AMMLP) algorithm gives priority to updating the weights for the less frequent activations over the more frequent ones. In this way metaplasticity is modeled artificially. AMMLP achieves a more effcient training, while maintaining MLP performance. To test the proposed algorithm we used the Wisconsin Breast Cancer Database (WBCD). AMMLP performance is tested using classification accuracy, sensitivity and specificity analysis, and confusion matrix. The obtained AMMLP classification accuracy of 99.26%, a very promising result compared to the Backpropagation Algorithm (BPA) and recent classification techniques applied to the same database.

In this communication, we propose the use of Support Vector Machines (SVM) for crop classification using hyperspectral images. SVM are benchmarked to well-known neural networks such as multilayer perceptrons (MLP), Radial Basis Functions (RBF) and Co-Active Neural Fuzzy Inference Systems (CANFIS). Models are analyzed in terms of efficiency and robustness, which is tested according to their suitability to real-time working conditions whenever a preprocessing stage is not possible. This can be simulated by considering models with and without a preprocessing stage. Four scenarios (128, 6, 3 and 2 bands) are thus evaluated. Several conclusions are drawn: (1) SVM yield better outcomes than neural networks; (2) training neural models is unfeasible when working with high dimensional input spaces and (3) SVM perform similarly in the four classification scenarios, which indicates that noisy bands are successfully detected.

Abdominal aortic aneurysm (AAA) is a condition where the weakening of the aortic wall leads to its widening and the generation of a thrombus. To prevent a possible rupture of the aortic wall, AAA can be treated non-invasively by means of the endovascular aneurysm repair technique (EVAR), consisting of placing a stent-graft inside the aorta by a cateter to exclude the aneurysm sac from the blood circulation. A major complication is the presence of liquid blood turbulences, called endoleaks, in the thrombus formed in the space between the aortic wall and the stent-graft. In this paper we propose an automatic method for the detection of type II endoleaks in computer tomography angiography (CTA) images. The lumen and thrombus in the aneurysm area are first segmented using a radial model approach. Then, these regions are split into Thrombus Connected Components (TCCs) using a watershed-based segmentation and geometric and image content-based characteristics are obtained for each TCC. Finally, TCCs are classified into endoleaks and non-endoleaks using a multilayer Perceptron (MLP) trained on manual labeled sample TCCs provided by experts.

network classifiers have been widely used in classification due to its adaptive and parallel processing ability. This paper concerns classification of underwater passive sonar signals radiated by ships using neural networks. Classification process can be divided into two stages: one is the signa preprocessing and feature extraction, the other is the recognition process. In the preprocessing and feature extraction stage, the wavelet transform (WT) is used to extract tonal features from the average power spectral density (APSD) of the input data. In the classification stage, two kinds of neural network classifiers are used to evaluate the classification results, inclusive of the hyperplanebased classifier-Multilayer Perceptron (MLP)-and the kernel-based classifier-Adaptive Kernel Classifier (AKC). The experimental results obtained from MLP with different configurations and algorithms show that the bipolar continuous function possesses a wider range and a higher value of the learning rate than the unipolar continuous function. Besides, AKC with fixed radius (modified AKC) sometimes gives better performance than AKC, but the former takes more training time in selecting the width of the receptive field. More important, networks trained with tonal features extracted by WT has 96% or 94% correction rate, but the training with original APSDs only have 80% correction rate. &words-Underwater signal classification, Neural networks, Wavelet transform, Multilayer perceptron, Adaptive kernel classifier.

A reconfiguration technique for multilevel inverters incorporating a diagnostic system based on neural network is proposed in this paper. It is difficult to diagnose a multilevel-inverter drive (MLID) system using a mathematical model because MLID systems consist of many switching devices and their system complexity has a nonlinear factor. Therefore, a neural network (NN) classification is applied to the fault diagnosis of a MLID system. Multilayer perceptron networks are used to identify the type and location of occurring faults. The principal component analysis (PCA) is utilized in the feature extraction process to reduce the NN input size. A lower dimensional input space will also usually reduce the time necessary to train a NN, and the reduced noise may improve the mapping performance. The output phase voltage of a MLID can be used to diagnose the faults and their locations. The reconfiguration technique is also proposed. The effects of using the proposed reconfiguration technique at high modulation index are addressed. The proposed system is validated with experimental results. The experimental results show that the proposed system performs satisfactorily to detect the fault type, fault location, and reconfiguration.

The paper presents some contemporary approaches to spatial environmental data analysis. The main topics are concentrated on the decision-oriented problems of environmental spatial data mining and modeling: valorization and representativity of data with the help of exploratory data analysis, spatial predictions, probabilistic and risk mapping, development and application of conditional stochastic simulation models. The innovative part of the paper presents integrated/hybrid model-machine learning (ML) residuals sequential simulations-MLRSS. The models are based on multilayer perceptron and support vector regression ML algorithms used for modeling long-range spatial trends and sequential simulations of the residuals. ML algorithms deliver non-linear solution for the spatial non-stationary problems, which are difficult for geostatistical approach. Geostatistical tools (variography) are used to characterize performance of ML algorithms, by analyzing quality and quantity of the spatially structured information extracted from data with ML algorithms. Sequential simulations provide efficient assessment of uncertainty and spatial variability. Case study from the Chernobyl fallouts illustrates the performance of the proposed model. It is shown that probability mapping, provided by the combination of ML data driven and geostatistical model based approaches, can be efficiently used in decision-making process. 

An arti®cial neural network-based system is proposed to predict mechanical properties in spheroidal cast iron. Several castings of various compositions and modules were produced, starting from different inoculation temperatures and with different cooling times. The mechanical properties were then evaluated by means of tension tests. Process parameters and mechanical properties were then used as a training set for an arti®cial neural network. Different neural structures were tested, from the simple perceptron up to the multilayer perceptron with two hidden layers, and evaluated by means of a validation set. The results have shown excellent predictive capability of the neural networks as regards maximum tensile strength, when the variation range of strength does not exceed 100 MPa. 5 2000 Elsevier Science S.A. All rights reserved.

Electricity is indispensable and of strategic importance to national economies. Consequently, electric utilities make an effort to balance power generation and demand in order to offer a good service at a competitive price. For this purpose, these utilities need electric load forecasts to be as accurate as possible. However, electric load depends on many factors (day of the week, month of the year, etc.), which makes load forecasting quite a complex process requiring something other than statistical methods. This study presents an electric load forecast architectural model based on an Artificial Neural Network (ANN) that performs Short-Term Load Forecasting (STLF). In this study, we present the excellent results obtained, and highlight the simplicity of the proposed model. Load forecasting was performed in a geographic location of the size of a potential microgrid, as microgrids appear to be the future of electric power supply.

We present a new approach for classifying mpeg-2 video sequences as 'cartoon' or 'non-cartoon' by analyzing specific video and audio features of consecutive frames in real-time. This is part of the well-known video-genre- classification problem, where popular TV-broadcast genres like cartoon, commercial, music, news and sports are studied. Such applications have also been discussed in the context of MPEG-7 (12).

In this work, a channel estimation technique based on Artificial Neural Networks (ANN) has been proposed as an alternative to pilot based channel estimation technique for Space-Time Block Coded Multiple-Input Multiple-Output (STBC-MIMO) systems over Rayleigh fading channels. ANNs, due to their high degree of adaptability, can be used for modelling the nonlinear phenomenon of channel estimation and for decoding the degraded symbols over severely faded channel. Two different ANN structures, namely Multilayer Perceptron (MLP) and Recurrent Neural Networks (RNN) have been trained and tested extensively for estimating the channel in STBC-MIMO systems with Binary Phase Shift Keying (BPSK) modulation scheme. Simulated results in terms of Bit Error Rates (BER) vs. Signal to Noise Ratio (SNR) have been used to compare the effectiveness of the learning capability of the two ANN structures over wireless fading channel. RNN structures are found to outperform the MLP, both in terms of training performance and BER results and thus proving its nonlinear dynamic adaptive behaviour.

Induction motors are subject to different faults which, if undetected, may lead to serious machine failures. From the scrupulous review of related works, it is observed that neuro-fuzzy and neural network (NN)-based fault-detection schemes are performed well for large machines and they are not only expensive but also complex. In this paper, the authors developed the radial-basis-function-multilayer-perceptron cascadeconnection NN-based fault-detection scheme for the small and medium sizes of three-phase induction motors. Stator winding interturn short, rotor eccentricity, and both faults simultaneously are selected for demonstration. Simple statistical parameters of stator current are considered as input features. Principal component analysis is used to select suitable inputs to the network. Experimental results are included to show the ability of the proposed classifier for detecting faults. Moreover, the network is tested for the robustness to the uniform and Gaussian noises. Having good classification accuracy with enough robustness to noises, the proposed classifier is suitable for the real-world applications.

In this paper, we solve the customer credit card churn prediction via data mining. We developed an ensemble system incorporating majority voting and involving Multilayer Perceptron (MLP), Logistic Regression (LR), decision trees (J48), Random Forest (RF), Radial Basis Function (RBF) network and Support Vector Machine (SVM) as the constituents. The dataset was taken from the Business Intelligence Cup organised by the University of Chile in 2004. Since it is a highly unbalanced dataset with 93% loyal and 7% churned customers, we employed (1) undersampling, (2) oversampling, (3) a combination of undersampling and oversampling and (4) the Synthetic Minority Oversampling Technique (SMOTE) for balancing it. Furthermore, tenfold cross-validation was employed. The results indicated that SMOTE achieved good overall accuracy. Also, SMOTE and a combination of undersampling and oversampling improved the sensitivity and overall accuracy in majority voting. In addition, the Classification and Regression Tree (CART) was used for the purpose of feature selection. The reduced feature set was fed to the classifiers mentioned above. Thus, this paper outlines the most important predictor variables in solving the credit card churn prediction problem. Moreover, the rules generated by decision tree J48 act as an early warning expert system.

This paper presents the use of a neural network and a decision tree, which is evolved by genetic programming (GP), in thalassaemia classification. The aim is to differentiate between thalassaemic patients, persons with thalassaemia trait and normal subjects by inspecting characteristics of red blood cells, reticulocytes and platelets. A structured representation on genetic algorithms for non-linear function fitting or STROGANOFF is the chosen architecture for genetic programming implementation. For comparison, multilayer perceptrons are explored in classification via a neural network. The classification results indicate that the performance of the GP-based decision tree is approximately equal to that of the multilayer perceptron with one hidden layer. But the multilayer perceptron with two hidden layers, which is proven to have the most suitable architecture among networks with different number of hidden layers, outperforms the GP-based decision tree. Nonetheless, the structure of the decision tree reveals that some input features have no effects on the classification performance. The results confirm that the classification accuracy of the multilayer perceptron with two hidden layers can still be maintained after the removal of the redundant input features. Detailed analysis of the classification errors of the multilayer perceptron with two hidden layers, in which a reduced feature set is used as the network input, is also included. The analysis reveals that the classification ambiguity and misclassification among persons with minor thalassaemia trait and normal subjects is the main cause of classification errors. These results suggest that a combination of a multilayer perceptron with a blood cell analysis may give rise to a guideline/hint for further investigation of thalassaemia classification.

This study investigated whether two artificial neural networks (ANNs), multilayer perceptron (MLP) and hybrid networks using statistical and ANN approaches, can outperform traditional statistical models for predicting corporate failures in Australia one year and two years prior to the financial distress. The results suggest that hybrid neural networks outperform all other models one and two years before failure. Therefore, hybrid neural network model is a very promising tool for failure prediction. This supports the conclusion that for shareholders, policymakers and others interested in early warning systems, hybrid networks would be useful.

In this work, a channel estimation technique based on Artificial Neural Networks (ANN) has been proposed as an alternative to pilot based channel estimation technique for Space-Time Block Coded Multiple-Input Multiple-Output (STBC- MIMO) systems over Rayleigh fading channels. ANNs, due to their high degree of adaptability, can be used for modelling the nonlinear phenomenon of channel estimation and for decoding the degraded symbols over severely faded channel. Two different ANN structures, namely Multilayer Perceptron (MLP) and Recurrent Neural Networks (RNN) have been trained and tested extensively for estimating the channel in STBC- MIMO systems with Binary Phase Shift Keying (BPSK) modulation scheme. Simulated results in terms of Bit Error Rates (BER) vs. Signal to Noise Ratio (SNR) have been used to compare the effectiveness of the learning capability of the two ANN structures over wireless fading channel. RNN structures are found to outperform the MLP, both in terms of training...

A pressure ulcer is a clinical pathology of localized damage to the skin and underlying tissue caused by pressure, shear, or friction. Diagnosis, treatment, and care of pressure ulcers are costly for health services. Accurate wound evaluation is a critical task for optimizing the efficacy of treatment and care. Clinicians usually evaluate each pressure ulcer by visual inspection of the damaged tissues, which is an imprecise manner of assessing the wound state. Current computer vision approaches do not offer a global solution to this particular problem. In this paper, a hybrid approach based on neural networks and Bayesian classifiers is used in the design of a computational system for automatic tissue identification in wound images. A mean shift procedure and a region-growing strategy are implemented for effective region segmentation. Color and texture features are extracted from these segmented regions. A set of multilayer perceptrons is trained with inputs consisting of color and texture patterns, and outputs consisting of categorical tissue classes which are determined by clinical experts. This training procedure is driven by a -fold cross-validation method. Finally, a Bayesian committee machine is formed by training a Bayesian classifier to combine the classifications of the neural networks. Specific heuristics based on the wound topology are designed to significantly improve the results of the classification. We obtain high efficiency rates from a binary cascade approach for tissue identification. Results are compared with other similar machine-learning approaches, including multiclass Bayesian committee machine classifiers and support vector machines. The different techniques analyzed in this paper show high global classification accuracy rates. Our binary cascade approach gives high global performance rates (average sensitivity = 78 7%, specificity = 94 7%, and accuracy = 91 5%) and shows the highest average sensitivity score (=86.3%) when detecting necrotic tissue in the wound.

In this work a novel method for human face recognition that is based on fuzzy neural network has been presented. Here, Gabor wavelet transformation is used for extraction of features from face images as it deals with images in spatial as well as in frequency domain to capture different local orientations and scales efficiently. In face recognition problem multilayer perceptron (MLP) has already been adopted owing to its efficiency, but it does not capture overlapping and nonlinear manifolds of faces which exhibit different variations in illumination, expression, pose, etc. A fuzzy MLP on the other hand performs better than an MLP because fuzzy MLP can identify decision surfaces in case of nonlinear overlapping classes, whereas an MLP is restricted to crisp boundaries only. In the present work, a new approach for fuzzification of the feature sets obtained through Gabor wavelet transforms has been discussed. The feature vectors thus obtained are classified using a newly designed fuzzified MLP. The system has been tested on a composite database (DB-C) consisting of the ORL face database and another face database created for this purpose and a recognition rate of 97.875% with fuzzy MLP against a recognition rate of only 81.25% with MLP whose feature vectors were also obtained through same Gabor wavelet transforms has been obtained.

This paper presents a new hybrid optimization strategy for training feedforward neural networks. The algorithm combines gradient-based optimization of nonlinear weights with singular value decomposition (SVD) computation of linear weights in one integrated routine. It is described for the multilayer perceptron (MLP) and radial basis function (RBF) networks and then extended to the local model network (LMN), a new feedforward structure in which a global nonlinear model is constructed from a set of locally valid submodels. Simulation results are presented demonstrating the superiority of the new hybrid training scheme compared to second-order gradient methods. It is particularly effective for the LMN architecture where the linear to nonlinear parameter ratio is large.

In this study, fluidized-bed manufactured enteric-coated omeprazole pellets were compressed into tablets. The stability of the pellets and those of compressed tablets were evaluated for remaining omeprazole and for degradation products under an accelerated stability protocol. The data were analyzed using the artificial neural network (ANN) and analysis of variance (ANOVA). It was found that enteric-coated omeprazole pellets could be compressed into quickly disintegrating tablets using microcrystalline cellulose granules as the pressure absorbing matrix. The ANN, using the multilayer perceptron model, predicted that there was a positive correlation between tablet crushing strength and microcrystalline cellulose concentration. Microcrystalline cellulose matrix showed a strong plastic deformation and all the pellets inside the tablet maintained their integrity with no significant change in their surface properties. Omeprazole degradation in acid medium was mainly dependent on microcrystalline cellulose concentration. A 90-day accelerated stability test in brown glass bottles with a desiccant showed that all prototype formulations would result in an acceptable stability profile for both remaining omeprazole, and also for the increase of impurity concentrations. q

—Early detection of cancer is the most promising way to enhance a patient's chance for survival. This paper presents a computer-aided classification method using computed tomography (CT) images of the lung based on ensemble of three classifiers including MLP, KNN and SVM. In this study, the entire lung is first segmented from the CT images and specific features like Roundness, Circularity, Compactness, Ellipticity, and Eccentricity are calculated from the segmented images. These morphological features are used for classification process in a way that each classifier makes its own decision. Finally, majority voting method is used to combine decisions of this ensemble system. The performance of this system is evaluated using 60 CT scans collected by Lung Image Database Consortium (LIDC) and the results show good improvement in diagnosing of pulmonary nodules.

Churn represents the problem of losing a customer to another business competitor which leads to serious profit loss. Therefore, many companies investigate different techniques that can predict churn rates and help in designing effective plans for customer retention. In this study we investigate the application of Multilayer Perceptron (MLP) neural networks with back-propagation learning for churn prediction in a telecommunication company. Different MLP topologies with different settings are used to build churn classification models. Moreover, two MLP based approaches are used and compared in order to rank the most influencing factors on churn rates. For the purpose of this research, real data of customers in a major Jordanian telecom company were provided. [Omar Adwan, Hossam Faris, Khalid Jaradat, Osama Harfoushi and Nazeeh Ghatasheh. Predicting Customer Churn in Telecom Industry using MLP Neural Networks: Modeling and Analysis. Life Sci J 2014;11(3):75-81]. (ISSN:1097-8135). http:...

Neural networks are one of the widely-used time series forecasting methods in time series applications. Among different neural network architectures and learning algorithms, the most popular choice is the feedforward Multilayer Perceptron (MLP). However, it suffers from some drawbacks such as getting trapped in local minima, human intervention during the stage of training, and limitations in architecture design. The aims of this study were twofold. The first was to employ NeuroEvolution of Augmenting Topologies (NEAT), which has many successful applications in numerous fields. In this paper, we applied it to time series forecasting for the first time and compared its performance with that of the MLP. The second aim was to analyse the performance resulting from the pairwise combination of these methods. In general, the results suggested that the forecasts from the NEAT algorithm were more accurate than those of the MLP. The results also showed that pairwise combined forecasts in gene...

Complex jargon represents an impediment to newcomers to the field of neural networks. This document presents a glossary of some of the basic terminology that may be encountered in the literature, with an emphasis on the use of multilayer perceptrons and radial basis functions for regression problems.