Feedforward Fuzzy Neural Network Research Articles

Max-C and Min-D Projection Auto-Associative Fuzzy Morphological Memories: Theory and an Application for Face Recognition

Max-C and min-D projection auto-associative fuzzy morphological memories (max-C and min-D PAFMMs) are two-layer feedforward fuzzy morphological neural networks designed to store and retrieve finite fuzzy sets. This paper addresses the main features of these auto-associative memories: unlimited absolute storage capacity, fast retrieval of stored items, few spurious memories, and excellent tolerance to either dilative or erosive noise. Particular attention is given to the so-called Zadeh’ PAFMM, which exhibits the most significant noise tolerance among the max-C and min-D PAFMMs besides performing no floating-point arithmetic operations. Computational experiments reveal that Zadeh’s max-C PFAMM, combined with a noise masking strategy, yields a fast and robust classifier with a strong potential for face recognition tasks.

Efficient Algorithm for Solving Fuzzy Singularly Perturbed Volterra Integro-Differential Equation

In this paper, we design a fuzzy neural network to solve fuzzy singularly perturbed Volterra integro-differential equation by using a High Performance Training Algorithm such as the Levenberge-Marqaurdt (TrianLM) and the sigmoid function of the hidden units which is the hyperbolic tangent activation function. A fuzzy trial solution to fuzzy singularly perturbed Volterra integro-differential equation is written as a sum of two components. The first component meets the fuzzy requirements, however, it does not have any fuzzy adjustable parameters. The second component is a feed-forward fuzzy neural network with fuzzy adjustable parameters. The proposed method is compared with the analytical solutions. We find that the proposed method has excellent accuracy in findings, a lower error rate, and faster convergence than other typical methods.

Fully Fuzzy Neural Network For Solving Fuzzy Differential Equations

In this work we have introduced a new method for solving fuzzy differential equations .This method "based on the" fully fuzzy "neural network" to find "the numerical solution of" the "first order fuzzy differential equations" ."The fuzzy trial solution of" the "fuzzy initial value problem is written as a sum of two parts". The first part satisfies the fuzzy condition, it contains no fuzzy adjustable parameters. The second part involves fully fuzzy feed-forward neural networks containing fuzzy adjustable parameters. "Under some conditions the proposed method provides numerical solutions with high accuracy".

Using Fuzzy Feed-Forward Neural Network for Linguistic Processing

Fuzzy sets have been implemented efficiently to manage unclear data, language terms, andvague notions. Recently, considerable work has been dedicated to merging neural-networktechniques with fuzzy sets. In this study, present the structure of a fuzzy feed-forward neuralnetwork (FFFNN) with a trapezoidal fuzzy set. In addition to handling real input vectors, it isalso capable of handling fuzzy input vectors. Generally, the output of a FNN is a fuzzy vector.According to the extension principle of Zadeh, each unit of a FNN has an input-outputrelationship. To determine the costs associated with fuzzy calculations and fuzzy objectives,developed a cost function. At that point, created a learning algorithm from the cost capacity toalign the four variables of each trapezoidal fuzzy weight. In conclusion, demonstrate ourmethodology using numerical models.

Fuzzy fractional more sigmoid function activated neural network approximations revisited

<p style='text-indent:20px;'>Here we study the univariate fuzzy fractional quantitative approximation of fuzzy real valued functions on a compact interval by quasi-interpolation arctangent-algebraic-Gudermannian-generalized symmetrical activation function relied fuzzy neural network operators. These approximations are derived by establishing fuzzy Jackson type inequalities involving the fuzzy moduli of continuity of the right and left Caputo fuzzy fractional derivatives of the involved function. The approximations are fuzzy pointwise and fuzzy uniform. The related feed-forward fuzzy neural networks are with one hidden layer. We study also the fuzzy integer derivative and just fuzzy continuous cases. Our fuzzy fractional approximation result using higher order fuzzy differentiation converges better than in the fuzzy just continuous case.</p>

A novel hybridized neuro-fuzzy model for solving fuzzy singular perturbation problems with initial conditions

The present paper tries to introduce a new process for solving fuzzy singular perturbation problem(SPP, s) with initial condition. This approach depends on on the partially fuzzy neural network to find the numerical solution of the second order of these problems. This system’s trial solution is written as a sum of two parts. The first section meets the fuzzy initial condition and does not have any fuzzy free parameters. The second component consists of a partially fuzzy feed-forward neural network. containing fuzzy adjustable parameters (the fuzzy weights). As a result, the starting condition is fulfilled by construction, and the network is trained to solve the differential equations. When compared to other numerical techniques, this technique proves that neural networks generate solutions with high generalizability and accurateness. A number of examples are given to show the proposed plan.

Neuro-fuzzy system for solving fuzzy singular perturbation problems

In this paper we have introduced a new method, by using fully fuzzy neural network (FFNN) to find the numerical solution for solving fuzzy singular perturbation problems (FSPPS) for ordinary differential equations. It uses the Levenberg-Marquardt training machine algorithm (TrianLM) for calculating the Fuzzy neural network parameters and the sigmoid function of the hidden units is hyperbolic tangent activation function, that is: . The fuzzy trial solution of FSPPs is written as a sum of two parts. The first part satisfies the fuzzy conditions, it contains no fuzzy adjustable parameters. The second part involves a feed-forward fuzzy neural network containing fuzzy adjustable parameters. In comparison with existing numerical methods, shows upp that the use of fuzzy neural networks provides solutions with good generalization and high accuracy. The proposed method is illustrated by several examples.

Constructing a Precise Fuzzy Feedforward Neural Network Using an Independent Fuzzification Approach

This study discusses how to fuzzify a feedforward neural network (FNN) to generate a fuzzy forecast that contains the actual value, while minimizing the average range of fuzzy forecasts. This topic has rarely been investigated in past studies, but is an essential step to constructing a precise fuzzy FNN (FFNN). Existing methods fuzzify all parameters at the same time, which re-sults in a nonlinear programming (NLP) problem that is not easy to solve. In contrast, in this study, the parameters of a FNN are fuzzified independently. In this way, the optimal values of fuzzy parameters can be derived theoretically. An illustrative example is used to illustrate the ap-plicability of the proposed methodology. According to the experimental results, fuzzifying the thresholds on hidden-layer nodes or the connection weights between input and hidden layers may not guarantee that all fuzzy forecasts contain the corresponding actual values. In contrast, fuzzi-fying the threshold on the output node and the connection weights between the hidden and out-put layers is more likely to achieve a 100% hit rate. The results lay a foundation for establishing a precise deep FFNN in the future.

Fuzzy feed forward neural network (FFFNN) model for the Jakarta Islamic index (JII) forecasting

Feed Forward Neural Network (FFNN) model is the best model to forecast the time series data. In this research, The Fuzzy Feed Forward Neural Network (FFFNN) with backward propagation method is used to predict the Jakarta Islamic Index (JII) data time series in 2018. Fuzzy is used as input to the FFNN model because it is overcome the weaknesses of the inaccurate results of the FFNN when the data is unclear or incomplete. The purposes of this research are to explain the procedure to generate the FFFNN model. The steps of the FFFNN model prepare the input data to become a fuzzy number using Growth-S Curve fuzzification; the second is divide the data into two training and testing data, the third is determining the best neural network architecture with different neurons and hidden layers to get the best weights that used for the forecasting model. In this research, the best FFFNN model is built by 19 neurons and one hidden layer with 90% and 10% training and testing data, respectively. Therefore with the model obtained, forecasting produces the value of MSE 0.0018 in training and 0.0004 in testing. From the MSE values obtained, it can be concluded that the forecasting using FFFNN model is reasonable to predict.

A fast learning algorithm based on extreme learning machine for regular fuzzy neural network

The regular fuzzy neural network (RFNN) is a kind of fuzzy neural network by fuzzifying the feed-forward neural network. The RFNN can directly deal with the language information and it has the merits of fuzzy system and neural network. It is presented a fast learning algorithm based on the extreme learning machine (ELM) for the RFNN in this paper. The RFNN referred here is a three-layer feed-forward fuzzy neural network and the connected weights in the RFNN are all fuzzy numbers. A simulation example is given to approximately realize the fuzzy if-then rules by the RFNN. The results show that the RFNN trained by the proposed algorithm has good performance and approximation ability.

Precise control and prediction of the greenhouse growth environment of Dendrobium candidum

ObjectiveDendrobium officinale is a perennial epiphytic herb of the Orchidaceae family. It has very strict needs with respect to growth environment and climatic conditions. In order to precisely control the artificial growth environment of Dendrobium candidum, a monitoring system was designed and implemented using a Programmable Logic Controller (PLC) and Supervisory Control and Data Acquisition (SCADA) system. MethodsA platform was created for precise control using a Siemens S7-200CPU224 PLC and different sensors of various environmental parameters as the control unit and SCADA-configured software as the core of the monitoring unit. We here used 4 indexes, soil temperature, soil moisture, humidity, and light, as prediction parameters and established a three-layer feed-forward fuzzy optimization neural network model. ResultsThe system not only allows prediction of the optimal environmental parameters for the growth of Dendrobium candidum, real-time monitoring, and intelligent control but also escapes the shortcomings of traditional back-propagation (BP) neural networks, which suffer from slow convergence, shock, and poor generalization. The current model’s average prediction error is less than 2.5%. It also provides a theoretical basis and decision support for the precision control of planting projects and relevant environment forecasting. The climate in the test area is hot and rainy in summer and colder and drier in winter. The annual precipitation is concentrated in spring and summer, peaking twice, in May and October. The subtropical high temperature was recorded in August, which has little rainfall and is prone to drought. Winter features both cold and warm air and some rainy days, but not as much overall precipitation as summer.

Chebyshev polynomial functions based locally recurrent neuro-fuzzy information system for prediction of financial and energy market data

In this paper Chebyshev polynomial functions based locally recurrent neuro-fuzzy information system is presented for the prediction and analysis of financial and electrical energy market data. The normally used TSK-type feedforward fuzzy neural network is unable to take the full advantage of the use of the linear fuzzy rule base in accurate input–output mapping and hence the consequent part of the rule base is made nonlinear using polynomial or arithmetic basis functions. Further the Chebyshev polynomial functions provide an expanded nonlinear transformation to the input space thereby increasing its dimension for capturing the nonlinearities and chaotic variations in financial or energy market data streams. Also the locally recurrent neuro-fuzzy information system (LRNFIS) includes feedback loops both at the firing strength layer and the output layer to allow signal flow both in forward and backward directions, thereby making the LRNFIS mimic a dynamic system that provides fast convergence and accuracy in predicting time series fluctuations. Instead of using forward and backward least mean square (FBLMS) learning algorithm, an improved Firefly-Harmony search (IFFHS) learning algorithm is used to estimate the parameters of the consequent part and feedback loop parameters for better stability and convergence. Several real world financial and energy market time series databases are used for performance validation of the proposed LRNFIS model.

Neural network and, fuzzy logic: an application to fingerprint recognition

In this paper, we utilize a feedforward fuzzy neural network to solve the problem of fingerprint-images classification into four classes: Whorl, Left Loop, Right Loop and Arch. At first, each fingerprint image is reduced noise, removed the background and then binarized. Then, one of thinning processes is chosen to perform on the image. Consequently, the 6x5 directional matrix which represents the global flow shapes of the fingerprint is established. After being transformed into 6 x 5 directional matrices, the four matrices of the above classes will be utilized as training patterns for a ANN. After learning from the four training pasterns above, our trial FNN has been experimented through 53 fingerprint samples and results in the classification rate of 96.23%. The obtained testing results demonstrate the system's power of classification.

Numerical Solution of Fuzzy Partial Differential Equations by Using Modified Fuzzy Neural Networks

The aim of this work is to present a modified method for finding the numerical solutions of fuzzy partial differential equations by using fuzzy artificial neural networks. Using a fuzzy trial neural solution depending on the fuzzy initial values and the fuzzy boundary conditions of the problem. Using modified fuzzy neural network makes that training points should be selected over an open interval without training the network in the range of first and end points. In fact, This new method based on replacing each element in the training set by a polynomial of first degree. The fuzzy trial solution of fuzzy partial differential equation is written as a sum of two parts. The first part satisfies the fuzzy conditions, it contains no fuzzy adjustable parameters. The second part involves a feed-forward fuzzy neural network containing fuzzy adjustable parameters. In comparison with existing similar fuzzy neural networks, the proposed method provides solutions with high accuracy. Finally, we illustrate our approach by two problems.

Numerical Solution of Fuzzy Differential Equations Based on Taylor Series by Using Fuzzy Neural Networks

In this paper a new method based on learning algorithm of Fuzzy neural network and Taylor series has been developed for obtaining numerical solution of fuzzy differential equations.A fuzzy trial solution of the fuzzy initial value problem is written as a sum of two parts.The first part satisfies the fuzzy initial condition,it contains Taylor series and involves no fuzzy adjustable parameters.The second part involves a feed-forward fuzzy neural network containing fuzzy adjustable parameters (the fuzzy weights).Hence by construction,the fuzzy initial condition is satisfied and the fuzzy network is trained to satisfy the fuzzy differential equation . In comparison with existing similar neural networks,the proposed method provides solutions with high accuracy.Finally , we illustrate our approach by two numerical examples .

NEURO-FUZZY MODELLING IN ANAEROBIC WASTEWATER TREATMENT FOR PREDICTION AND CONTROL

The aim of the present paper is to develop neuro-fuzzy prediction models in MATLAB environment of the anaerobic organic digestion process in wastewater treatment from laboratory and simulated experiments accounting for the variable organic load, ambient influence and microorganisms state. The main contributions are determination of significant model parameters via graphical sensitivity analysis, simulation experimentation, design and study of two “black-box” models for the biogas production rate, based on classical feedforward backpropagation and Sugeno fuzzy logic neural networks respectively. The models application is demonstrated in process predictive control

The Hybrid Method of Fuzzy Feed-Forward Neural Network for Predicting Protein Secondary Structure

With respect to the fact that the prediction of Protein secondary structure based on amino acids is very important, therefore, this study tries to present a new method based on the fuzzy combinational structure of a set of feed-forward neural networks so that the prediction accuracy of Protein secondary structure can be improved compared with the existing methods. Neural networks used in this paper are based on time windows; also, different methods have been established and trained to infer the three states of alpha- helix, beta- sheet and coils from DSSP results, and finally, combining the results of the abovementioned networks in a fuzzy manner, the prediction method of Protein secondary structure based on neural network has been improved. It should be noted that in this paper, CB513 and RS126 data sets which are valid data sets in evaluating prediction methods of Protein secondary structure known in research studies in this area have been used to train and evaluate the proposed method.

Long-term load forecasting by a collaborative fuzzy-neural approach

Long-term power load forecasting is of major importance for power suppliers to define the future power consumption of a given region. However, it is not easy to contend with the uncertainty of the long-term load. In order to effectively forecast the long-term load, a collaborative principal component analysis and fuzzy feed-forward neural network (PCA-FFNN) approach is proposed in this study. The difference between this and existing methods is that the collaborative PCA-FFNN approach takes into account the different points of view in a more efficient way, and therefore the results obtained are more comprehensive and more in-depth. In the proposed methodology, a group of domain experts is formed. These domain experts are asked to configure their own PCA-FFNNs to forecast the long-term load based on their views. A collaboration mechanism is therefore established. To facilitate the collaboration process and to derive a single representative value from these forecasts, the partial-consensus fuzzy intersection and radial basis function network (PCFI-RBF) approach is used. The effectiveness of the proposed methodology is illustrated with a case study.

Forecasting the yield of a semiconductor product with a collaborative intelligence approach

Forecasting the yield of a semiconductor product is an important task to the manufacturer. However, it is not easy to deal with the uncertainty in the yield. In order to effectively forecast the yield, a collaborative intelligence approach is proposed in this study. The difference with the existing methods is that the collaborative intelligence approach takes into account the different points of view in a more efficient way, and therefore the results obtained are more comprehensive and more reliable. In the collaborative intelligence approach, a group of domain experts is formed. These domain experts are asked to configure their fuzzy feed-forward neural networks (FFNNs) to forecast the yield based on their views. A collaboration mechanism is therefore established to evolve the views. To facilitate the collaboration process and to derive a single representative value from these forecasts, the maximal-consensus and radial basis function network (MC-RBF) approach is used. The effectiveness of the proposed methodology is shown with a case study.

The Newly AND-OR FNN Modeling and Application

The novel multilayer feed-forward AND-OR fuzzy neural network (AND-OR FNN) is proposed in this paper. The main feature is shown not only in reducing the input space by special inner structure of neurons, but also auto-extracting the rules by the structure self-organization and parameter self-learning. The equivalent is proved that the network structure and fuzzy inference. The whole structure of network is optimized by genetic algorithm to extract if-then rules. This designing approach is employed to modeling an AND-OR FNN controller for ship control. Simulated results demonstrate that the number of rule base is decreased remarkably and the performance is much better than ordinary fuzzy control, illustrate the approach is practicable, simple and effective.