This work deals with the potential application of artificialneuralnetworks to model sunshineduration in three cities in Algeria using ten input parameters. These latter are: year and month, longitude, latitude and altitude of the site, minimum, mean and maximum air temperature, wind speed and relative humidity. They were selected according to their availability in meteorological stations and based on the fact that they are considered as the most used parameters by researchers to model sunshinedurationusingartificialneuralnetworks. Several network architectures were tested to choose the most accurate and simple scheme. The optimum number of layers and neurons was determined by trial and error method. The optimized network was obtained using Levenberg-Marquardt back-propagation algorithm, one hidden layer including 25 neurons with Tan-sigmoid transfer function. The model developed in this study has the ability to estimate sunshineduration with a mean absolute percentage error value equals to 2.015%, a percentage root mean square error of 2.741% and a determination coefficient of 0.9993 during test stage.
Weilin Li proposes a fault detection and classification method for medium voltage DC (MVDC) shipboard power systems (SPSs) by integrating wavelet transform (WT) multire solution analysis (MRA) technique with artificialneuralnetworks (ANNs). The MVDC system under consideration for future all-electric ships presents a range of new challenges, in particular the fault detection and classification issues addressed in this paper. The WT-MRA and Parseval's theorem are employed in this paper to extract the features of different faults. The energy variation of the fault signals at different resolution levels are chosen as the feature vectors. As a result of analysis and comparisons, the Daubechies 10 (db10) wavelet and scale 9 are the chosen wavelet function and decomposition level. Then, ANN is adopted to automatically classify the fault types according to the extracted features. Different fault types, such as short circuit faults on both dc bus and ac side, as well as ground fault, are analyzed and tested to verify the effectiveness of the proposed method. These faults are simulated in real time with a digital simulator and the data are then initially analyzed with MATLAB. The casestudy is a notional MVDC SPS model, and promising classification accuracy can be obtained according to simulation results. Finally, the proposed fault detection algorithm is implemented and tested on a real-time platform, which enables it for future practical use .
An artificialneural network (ANN) is a data processing method which has analogous performance to the biological neuralnetworks of the human brain (Haykin, 1999). MLPs are the most popular and the simplest type of ANN. These approaches are widely used to construct the relationship between input and outputs (Ahmed et al., 2015). Multi-layer perceptron are feed-forward networks which include one or more hidden layers (Haykin, 1999). The MLP applied in this research contained a three-layer architecture consisting of an input layer, a hidden layer and an output layer. Figure 2 shows a typical MLP feedforward network for this study with one hidden layer. According to Hornik et al. (1989), the advantages of MLPs make this method easy to use and capable of estimating any input/output relation for more accurate prediction. The Levenberg–Marquardt (LM) algorithm (Levenberg, 1944; Marquardt, 1963) is an efficient learning approach for multi-layer feedforward networks. This method is a modified version of the classic Newton approach for obtaining an optimum solution to the optimization problem. This method employs an approximation to the Hessian matrix in the following equation (Equation 1).
Depression – a word synonymous to today’s world. Every instance we come across various people who have suffered depression at some point of time in their life or suffering from it at this very moment. Is depression fully curable? Can medication fully recover a person of depression or the patient has still traces left in him/her? A big question which still stands not fully answered. Let’s try and find it. Neural Network one of the pillars of Artificial Intelligence is very similar to our central Nervous System (CNS) and depression affects our CNS very badly. Using these powerful tools of ArtificialNeural Network we can try to give a permanent cure to depression.
Lagaris, et al.  used artificialneuralnetworks (ANN) for solving ordinary differential equations and partial differential equations for both boundary value and initial value problems. Canh and Cong  presented a new technique for numerical calculation of viscoelastic flow based on the combination of neuralnetworks and Brownian dynamics simulation or stochastic simulation technique (SST). Hayati and Karami  used a modified neural network to solve the Berger’s equation in one-dimen- sional quasilinear partial differential equation.
Numerous different methods concerning filter tuning algorithms were presented in previous publications. In , filter tuning in time domain is shown. The described method requires a correctly tuned filter serving as a template and a skilled operator. In , a machine learning system is proposed, which employs techniques for pattern recognition and adaptive signal processing. In this method, a skilled operator is still indispensable. The computer-controlled method of tuning microwave filters proposed in . In this work, an approximate filter network model investigating the effects of input/output couplings is used. The automatic tuning for three-pole resonator filter is presented. The novelapproach, for filter tuning, is shown in paper . In this paper, an algorithm based on fuzzy logic is introduced, proving that such a method can be helpful in the identification of the tuning elements being the source of the detuning. Interesting tool ROBO- CAT, for automated tuning, was presented in . In this case the algorithms are based on coupling matrix extraction method, time domain and phase cloning. The new approach based on the direct search methods was presented in . It has been proved that the described methods are very effective, but require many screw changes, which is not recommended due to passive inter modulation (PIM) problems.
The parameters that were collected as the important parameters in stuck pipe at first steps were as follows: Mud properties which are Mud Weight (MW), Plastic Viscosity (PV), Yield Point (YP), 10-Second Gel Strength and 10-Minute Gel Strength (GL), Fluidloss (conventional API or High Pressure High Temperature (HPHT) API) and Solid content. Depths, bottom hole assembly size and length, drill pipe size, hole size, Rate Of Penetration (ROP), Annular velocity, pipe rotation in Revolutions Per Minute (RPM) and Mud cake thickness. Influence of the input parameters is considered as the key point for developing ANN models. Introducing more input parameters than required will result in a large network size and consequently decrease learning speed and efficiency . Since the drilling parameters that are involved in stuck pipe are numerous, it is essential to find the variables that are closely related to stuck pipe. In ANN studies, the following criteria have been mentioned . 1) There must be a spread of values of the parameter in the databases. This allows the neural network to more easily approximate the function.
Machine learning (ML) that automatically learns programs from data has become more and more popular for data analysis . Artificialneuralnetworks (ANNs), inspired by biological neuralnetworks, are one widely used class of ML algorithms and artificial intelligence techniques[93, 94]. It is an advanced interpolation method. ANNs represent human’s feeble attempts to replicate thinking or decision making processes in the brain (e.g., memory, recognition, prediction, and planning). ANNs are relatively crude electronic network models which try to mimic the neural structure of the human brain. The neurons in the brain responds to the new external stimuli based on previous experiences. It is natural proof that small energy efficient packages (connected neurons) are capable of solving some problems that are beyond the scope of current computers [95, 96]. Although ANNs have limited functionality compared with the human brain, this new method of problem solving provides a more graceful degradation during system overload than traditional methods. ANNs and other artificial intelligence approaches of computing are thought to be the next major advancement in the computing industry [95, 97].
There is no established way to determine in advance which ANN model configuration will perform best for a specific problem. Some pre-selection of inputs can be undertaken, for example by examination of linear correlations between each input and the target output. How- ever, as ANN models can accommodate non-linear relationships this procedure is not necessarily advantageous, and may eliminate useful inputs. The Infinity program is highly automated, sequentially testing hundreds of candidate neural network model configurations, and thousands of combinations of inputs, to select the optimised output. The Infinity program uses a pre-set formula incorporating RMSE, MAE and correlation coefficient (r) to evaluate the accuracy for each neural network model and a corresponding set of selected inputs tested. Based on this formula, the program determines which ANN model and set of inputs is optimal.
inefficient compared to the impressive speed of SpikeNet which overcomes this constraint. Therefore, a primary objective for the future is to provide an impulse and asynchronous version of the model developed by Giese and Poggio. For the subject of artificialneuralnetworks in digital image processing, we cannot make a panoramic citation for lifetime achievement on this topic. Alexandrina-elenapandelea,(2015) application of soft computing models on digital image has been considered to be an approach for a better result. The main objective of the present work is to provide a new approach for image recognition usingArtificialNeuralNetworks. Initially an original gray scale intensity image has been taken for transformation. The Input image has been added with Salt and Peeper noise. Adaptive median Filter has been applied on noisy image such that the noise can be removed and the output image would be considered as filtered Image. The estimated Error and average error of the values stored in filtered image matrix have been calculated with reference to the values stored in original data matrix for the purpose of checking of proper noise removal. Now for recognition, a new test image has been taken and the same steps as salt & pepper noise insertion, removal of noise using adaptive median filter as mentioned earlier have been applied to get a new test matrix. Now the average error of the second image with respect to original image has been calculated based on the both generated matrices 3.0 Image system design parameters and modeling System
Normally, waste tires are considered a serious pollution problem because the waste rubber is not easily bio-degradable even after a long period landfill treatment [1–7]. During the last decades, several tentative strategies have been conducted to investigate the potential reuse of the recycled waste-tires as an innovative technique and to recycle them in civil engi- neering application [1–8]. Therefore, the addition of rubber into concrete makes that concrete regarded as a lightweight material. The literature about the use of tire rubber particles in cement-based materials focuses on the use of tire rubber as an aggregate in concrete such as ground, crumb and chipped rubber. Mechanical properties of rubberized concretes depend on the type and the content of utilized rubber. It was revealed that adding rubber particles to cement-based material provide a lower compressive strength [1, 2, 9–11]. The decrease in the compressive strength can be justified by the well-estab- lished fact that the compressive strength of concrete depends on the aggregates and on their volumetric proportion. The understanding of the relationship between the macroscopic mechanical behavior and the microstructural properties (such as volume fraction) is far from satisfactory. Some research had already been done on soft computing techniques mainly artifi- cial neuralnetworks to identify a model and control of its dif- ferent components as: cement, fine aggregates, coarse aggre- gates, sand, admixtures, ground, crumb and chipped rubber. Artificialneuralnetworks (ANN) are known as intelligent methods for modeling the behavior of physical phenomena. ANN takes data samples rather than entire data sets to arrive at solutions, which saves both time and money.
Our results show that a neural network using binary vectors as input variables and five neuralnetworks as a single output can obtain the best results in diagnosing for some frequents types of headache. We can see that these results could offer new possibilities for the headache patients with access problems to the Health System. Despite the difficulties of epidemiological studies in headache, it is known that these diseases are highly prevalent, with large social costs (Boardman, 2003; Dahlöf and Solomon, 2006). Population studies vary widely, with some prevalence studies results showing over 90%. A recent study in southern Brazil shows annual prevalence of 80.8% (Queiroz, Barea and Blank, 2006). The ideal way to establish a diagnosis of headache is through consultation with a neurologist who uses structured diagnostic criteria, such as the International Classification of Headache Disorders. The specialist has the ability to identify not only the most common causes of headaches, but also to evaluate the doubtful and mixed cases, secondary headaches and rare cases (Marks and Rapoport, 1997). We know, however, that access to neurologist is limited and expensive, and impossible to many patients.
In an another study, which included extraction of proteins from colloidal suspension, Albert S. Kim and Huaiqun Chen(University of Hawaii)used an ArtificialNeural Network as the alternative approach to mathematical modeling. The aim of the study was to predict the long term flux decrease due to colloidal cake formation. Relation between flux and trans-membrane pressure, particle size, solution pH, ionic strength, and elapsed operation time during cross-flow MF/UF membrane filtration is studied.  Different conditions such as temperature, feed concentration, and axial velocity are kept constant for the experiments and also for ANN simulations. The emphasis of the work was, first, to reduce the amount of training data sets with a small network configuration in terms of the number of hidden layers and the number of neurons in each layer, and second, to predict new data sets that might not be available by giving the operation conditions belonging to the training data sets. To obtain the optimal network structure for prediction, various ANN network structures were , and the developed performance of the ANNs are calculated by estimating the difference between the predicted output and the target output in terms of root mean square error (RMSE).
ArtificialNeuralNetworks, also referred as NeuralNetworks (NN), neuro-computers, connectionist networks, parallel- distributed processors etc are intelligent systems inspired by biological neuron systems. ANN is being used to tackle wide gamut of problems in pattern recognition, clustering, function approximation, forecasting/prediction, optimization etc. ANN as models to perform function approximation has been found successful in various engineering problems and is being widely applied in Traffic Engineering problems. A typical Multi Forward Neural Network (MFNN) has three layers: the input layer, the hidden layer, and the output layer. Since NN has the capability of learning (Lee and Lee, 2003).The most popular and successful learning algorithm used to train MFNNs in areas such as pattern recognition, function fitting speech and natural language processing and system modelling is the Back Propagation (BP) algorithm. Standard back propagation is a gradient descent algorithm, in which the network weights are moved along the negative of the gradient of the performance function. Properly trained back propagation networks tend to give reasonable answers when presented with inputs that they have never seen. A typical NN architecture has been shown in Figure 1.
The encoder is based on a bank of phonological encoders realised as neural network classifiers, 3-hidden layer multilayer perceptrons (MLPs), that encode individual phonology features. Each MLP classifies a binary phonological feature. The French speech database Ester Galliano et al. (2006) of standard French radio broadcast news was used for training of the encoders. It comprised 120 speakers in various recording conditions. We hypothesised that the broadcast recordings are more suitable for “live” speech encoding. In this study, a subset of 112 hours of recordings was used for the training. The phoneme set comprising 38 phonemes (including “sil”) was defined by the BDLex Perennou (1986) lexicon.
and vi) support interconnection with other components. Taking these six electrical requirements into account, in this paper we build upon our previous work in [ 32 ] and develop a fully digital memristor hardware simulator based on a behavioral model of voltage-controlled threshold-type bipolar memristors [ 27 ]. The presented electronic module constitutes the core of a digital memristor emulator, which further requires interface circuitry to permit connection to external circuits as a two terminal element, such as in [ 27 ], [ 31 ]. We conducted all the required verification tests and validated its functionality using Altera Quartus II and ModelSim software (SW), targeting low-cost yet powerful field programmable gate array (FPGA) families. The FPGAs are reconfigurable electronic platforms well-suited to implement ANNs [ 33 ], [ 34 ] owing to their HW flexibility, which allows rapid prototyping of different ANN topologies and implementation strategies. In this context, we chose the FPGA as the target electronic platform and showed that our design is suitable for FPGA- based ANNs. Our motivation was to design and implement digital HW electronic synapses particularly based on memristive dynamics and prove their suitability and applicability to a variety of ANN-based applications. Compared to other emulation approaches from the recent literature, this digital design is compact, easily reconfigurable, demonstrates excellent matching with the memristor model on which it is based [ 27 ], and complies with all the aforementioned electrical requirements (i to vi). Moreover, we tested its suitability for anti-serial memristive interconnections [ 35 ] and proved its synapse functioning in single-layer perceptron, implementing examples of associative memory and a simplified variation of spike timing dependent plasticity (STDP) [ 36 ] unsupervised learning of spatio-temporal correlations in parallel input streams, following previous demonstrations in [ 31 ] and [ 37 ], respectively. We present the schematics of our digital circuit designs and comment on the required HW resource scaling, thereby providing a complete design framework for such memristor emulator-based ANNs.
Most colour applications of neuralnetworks are based on multi-layer perceptron feed-forward networks. These networks are described in detail by Shamey and Hussain but essentially map an input vector to an output vector via a hidden layer of processing units . The values of the weights (free parameters) in the network are determined by optimisation using a training set of input- output examples. Bishop et al. used a neural network to predict dye concentrations (for a three-dye system) from CIELAB coordinates . However, apart from in special cases, it is almost always better  to use the neural network to predict colour from recipes (analogous to the way in which Kubelka- Munk models, for example, operate) rather than attempting to predict recipes from colour directly. Westland used a neural network to predict spectral reflectance for mixtures of six printing inks printed on white card . A total of 123 samples were used to train the neural network and the performance of the network was then tested on 40 additional samples. The network outperformed a two-constant Kubelka-Munk model when the number of units in the hidden layer was 7 and when all of the available training data were used. It was shown that performance deteriorated as the number of training samples was reduced .
The face is the primary focus of attention and plays a major role in identification and establishing the uniqueness of a particular person from the rest of the human society. In spite of so many faces in the human society, there is remarkable ability of a human eye to recognized one face from another. A human can recognize thousands of faces learned throughout the lifetime and identify familiar faces at a glance even after years apart. This ability of human eye is quite effective, even though there are changes in the visual stimulus due to aging of a person, expression and change of looks due to glasses, beards or in hair style. There are many approaches such as security purpose, credit card verification, criminal identification etc. where the identification of a face plays an important role. A slight recognition of a particular person will be much better, in these kind of fields, than not even recognizing at all. Although it is clear that people are good at face recognition, but it’s not obvious that a human brain can encoded or decoded for every face. Human face recognition has been studied for more than twenty years. Developing a suitable program which can be used digitally in recognizing a face is a quite challenging task, because human faces are complex and are different from each other in every aspect. So, developing such a kind of program is a difficult task in this digital world, which may involve earlier techniques, which was used for recognition of faces, to make it reliable. For face identification the starting step involves extraction of the relevant features from facial images. A big challenge is how to quantize facial features so that a computer should be able to identify a face. The study carried out by many researchers over the past several years indicates that certain facial characteristics are used by human beings to identify faces.