Search

Nowadays we witness an increasing number of business frauds. To protect investors΄ interest, a financial firm should possess an effective means to detect such frauds. In this regard, artificial neural networks (ANNs) are widely used for fraud detection. Traditional back-propagation-based algorithms used for training an ANN, however, exhibit the local optima problem, thus reducing the effectiveness of an ANN in detecting frauds. To alleviate the problem, this paper proposes an approach to training an ANN using an ant colony optimization technique, through which the local optima problem can be solved and the effectiveness of an ANN in fraud detection can be improved. Based on our approach, an associated prototype system is designed and implemented, and an exploratory study is performed. The results of the study are encouraging, showing the viability of our proposed approach. [ABSTRACT FROM AUTHOR]

The paper presents design and evaluation of emotional state estimator based on artificial neural networks for physiology-driven adaptive virtual reality (VR) stimulation. Real-time emotional state estimation from physiological signals enables adapting the stimulations to the emotional response of each individual. Estimation is first evaluated on artificial subjects, which are convenient during software development and testing of physiology-driven adaptive VR stimulation. Artificial subjects are implemented in the form of parameterized skin conductance and heart rate generators that respond to emotional inputs. Emotional inputs are a temporal sequence of valence/arousal annotations, which quantitatively express emotion along unpleasant-pleasant and calm-aroused axes. Preliminary evaluation of emotional state estimation is also performed with a limited set of humans. Human physiological signals are acquired during simultaneous presentation of static pictures and sounds from valence/arousal-annotated International Affective Picture System and International Affective Digitized Sounds databases. [ABSTRACT FROM AUTHOR]

Dissolved Gas-in-oil Analysis (DGA) plays an important role in fault diagnosis of power transformers. BP (Back Propagation) algorithm is used to diagnosis for dissolves gases in the oil of transformer in this paper. But typical BP algorithm has some defects, such as converging slowly, searching space possessing local minima and oscillation. The algorithm using additional momentum method and L-M (Lerenberg-Marquardt) to train BP Neural Network has been proved to have good performance in avoiding the local trap and converging slowly. So this paper adopts BP artificial neural network with algorithm of additional momentum method and L-M in diagnosis of dissolves gases in the oil. A mass of gases samples are analyzed in the algorithm and the results are compared with the swatches forecasted. The comparison result indicates that the improved algorithm has better classify capability for single-gases swatch as well as high diagnosis precision. [ABSTRACT FROM AUTHOR]

Uranium extraction in Kazakhstan is carried out using underground leaching method. Economic performance depends on the production process speed and accuracy of geophysical data interpretation. Data interpretation can be performed using learned systems, such as artificial neural network (ANN), Linear Discriminant Analysis Classifier (LDAC), Support Vector Classification (SVM), k-Nearest-Neighbor (k-NN) and etc. In the paper ˵adjacency cube″ method for integration of results of few interpretation algorithms is proposed. Learning algorithm for the ˵adjacency cube″ with low computational complexity was developed. The proposed method improves quality of recognition by 2-3 percent. [ABSTRACT FROM AUTHOR]

Cooperative tasks to be successful in Collaborative Distributed Environments (CDE) require from users to be known. Before hard, which users are more suitable in the system to cooperate with, as well as which tools are needed to achieve the common goal in the system in a cooperative way. On the other hand, awareness allows users to be aware of others΄ activities each and every moment. Information about others΄ activities combined with their intentions and purposes could be used to improve cooperation in CDEs. This paper focuses on associating the concept of awareness with vector quantization techniques. It presents a novel non-supervised Artificial Neural Network (ANN) based model for CDE named CAwANN, which uses the information of awareness collaborations occurring in the environment in order to achieve the most suitable awareness-based collaboration / cooperation in the environment. [ABSTRACT FROM AUTHOR]

In this paper, a multi-branch back propagation neural network (BPNN) is adopted to predict the nonlinear seismic responses of an eccentric three-story reinforced concrete building. First of all, the network is trained in batch by the vibration table test data of the structure with the maximum acceleration of ground motion in 0.4g. Then, the trained network is used for structural responses prediction. The nonlinear structural acceleration responses of the each story are evaluated by the trained network for the maximum acceleration of ground motion in different amplitudes. Compared with the experimental results, it turns out that the trained network can accurately predict structural future dynamic responses. [ABSTRACT FROM AUTHOR]

A new approach for forecasting stock index based on Multi Expression Neural Network (MENN) is proposed in this paper. The approach employs the multi expression programming (MEP) to evolve the architecture of the MENN and the particle swarm optimization (PSO) to optimize the parameters encoded in the MENN. This framework allows input variables selection, over-layer connections for the various nodes involved. The performance and effectiveness of the proposed method are evaluated using stock market forecasting problems and compared with the related methods. [ABSTRACT FROM AUTHOR]

This paper proposes a new method for load forecasting–the wavelet neural network model for daily load forecasting. The neural call function is basis of nonlinear wavelets. A wavelet network is composed by the wavelet basis function. The global optimum solution is got. We overcome the intrinsic defects of a artificial neural network that its learning speed is slow, its network structure is difficult to determine rationally and it produces local minimum points. It can be seen from the example this method can improve effectively the forecast accuracy and speed. It can be applied to the daily agricultural electric load forecasting. [ABSTRACT FROM AUTHOR]

This paper presents a new variant of a well known competitive learning algorithm: Growing Hierarchical Recurrent Self Organizing Map (GH_RSOM). The proposed variant is like the basic Growing Hierarchical Self Organizing Map (GHSOM), however, in the GH_RSOM each map of each layer is a recurrent SOM (RSOM) it is characterized for each unit of the map by a difference vector which is used for selecting the best matching unit and also for adaptation of weights of the map. In this paper, we study the learning quality of the proposed GHSOM variant and we show that it is able to reach good vowels recognition rates. [ABSTRACT FROM AUTHOR]

Color information is of great importance for the tongue inspection of computer-aided tongue diagnosis system. However, the RGB signals generated by different imaging device varied greatly due to dissimilar lighting conditions and usage of different kinds of digital cameras. This is a key problem for the tongue inspection and diagnosis. A common solution is to correct the tongue images to standard color space by the aid of colorchecker. In this paper, three general color correction techniques: polynomial regression, artificial neural network and support vector regression (SVR) are applied to the color correction of tongue image and compared for their performance of accuracy and time complexity. The experimental results of colorchecker correction show that when properly optimized, SVR performs the best among these three algorithms, with a training error of 0 and a test error of 0.68 to 3.03. The polynomial regression algorithm performs a little worse, but it is more robust to the fluctuations of the environmental illuminant and much faster than SVR to train the parameters. The ANN performs worst, and it is also time-consuming to train. Performance comparison to correct real tongue images shows that polynomial regression is better than SVR to achieve a close correction result to human perception. Finally, this paper is concluded that for tongue inspection in a computer-aided tongue diagnosis system, polynomial regression is suitable for online system correction to aid tongue diagnosis, while SVR technique offer a better alternative for the offline and automated tongue diagnosis. [ABSTRACT FROM AUTHOR]

This paper explores recent works in the application of artificial neural network (ANN) for security – namely, network security via intrusion detection systems, and authentication systems. This paper highlights a variety of approaches that have been adopted in these two distinct areas of study. In the application of intrusion detection systems, ANN has been found to be more effective in detecting known attacks over rule-based system; however, only moderate success has been achieved in detecting unknown attacks. For authentication systems, the use of ANN has evolved considerably with hybrid models being developed in recent years. Hybrid ANN, combining different variants of ANN or combining ANN with non-AI techniques, has yielded encouraging results in lowering training time and increasing accuracy. Results suggest that the future of ANN in the deployment of a secure environment may lie in the development of hybrid models that are responsive for real-world applications. [ABSTRACT FROM AUTHOR]

This paper proposes a new technique using discrete wavelet transform (DWT) and artificial neural network for fault classifications on single circuit transmission lines. Simulations and the training process for the artificial neural network are performed using ATP/EMTP and MATLAB. The mother wavelet daubechies4 (db4) is employed to decompose, high frequency component from these signals. Positive sequence current signals are used in fault detection decision algorithm. The variations of first scale high frequency component that detect fault are used as an input for the training pattern. Back-propagation (BP) neural network is also compared with the RBF neural network in this paper. The result is shown that an average accuracy values obtained from RBF gives satisfactory results with less training time, and will be very useful in the development of a power system protection scheme. [ABSTRACT FROM AUTHOR]

This paper extends an inference test proposed in [1]. The seminal paper proposes an artificial neural network (ANN) based panel unit root test in a dynamic heterogeneous panel context. The ANN is not complex, but it is not necessarily in the aim of modeling macroeconomic time series. However, it is applied in a difficult mathematical context, in which the classical Gaussian asymptotic probabilistic theory does not apply. Some asymptotic properties for the test were set, however, the small sample properties are not satisfactory. Consequently, in this paper, we propose to use the simulation based numerical method named ˵bootstrap″ to compute the small sample distribution of the test statistics. An application to a panel of bilateral real exchange rate series with the US Dollar from the 20 major OECD countries is provided. [ABSTRACT FROM AUTHOR]

Optical Character Recognition is one of the rapidly growing areas of Artificial Intelligence due to its vast applicability. The technique is used to recognize characters printed on paper or elsewhere. The optical character recognition gains more importance when there are multiple languages present. The complexity of the problem increases for the addition of every language. The identification of character is both difficult and important in the presence of multiple languages. In this paper we propose a technique for multi lingual character identification. The algorithm uses the characteristics of the language to find out the language first. This is done using a rule-based approach. Then we apply neural network of the particular language to find out the exact character. We have coded and tested this using English capital letters, English small letters and Hindi letters. We got an appreciable accuracy using test cases of all languages. This proves the efficiency of the algorithm. [ABSTRACT FROM AUTHOR]

In this paper we present a control architecture for multi-robot systems in dynamic environments, where the low level behaviors are obtained through artificial neural networks and evolutionary algorithms to achieve collaborative behaviors in a multi-robot system. As an example, we have cooperative tasks establishing a surveillance scenario stressing cooperation and competition between them. [ABSTRACT FROM AUTHOR]

Various radio systems can be used to obtain the position information in indoor environments. Due to the ubiquitous presence of WLAN networks, positioning techniques in these environments are the scope of intense research. This paper explores the properties of cascade-connected Artificial Neural Networks (ANNs) structures. Several cascade-connected ANN structures with space partitioning are compared to the single ANN multilayer feedforward structure. The benefits of using cascade-connected ANNs structures are shown and discussed in terms of the size of the environment and subspaces. The optimal cascade-connected ANN structure with space partitioning shows a 41% decrease in median error with respect to the single ANN model. [ABSTRACT FROM AUTHOR]

This paper presents a combined approach to automatic extraction of blood vessels in retinal images. The proposed procedure is composed of two phases: a wavelet transform-based preprocessing phase and a NN-based one. Several neural net topologies and training algorithms are considered with the aim of selecting an effective combined method. Human retinal fundus images, derived from the publicly available ophthalmic database DRIVE, are processed to detect retinal vessels. The approach is tested by considering performances in terms of sensitivity and specificity values obtained from vessel classification. The quality of vessel identifications is evaluated on obtained image by computing both sensitivity values and specificity ones and by relating them in ROC curves. A comparison of performances by ROC curve areas for various methods is reported. [ABSTRACT FROM AUTHOR]

One of the applications of Data Mining is the extraction of knowledge from time series [1][2]. The Artificial Neural Networks (ANNs), one of the techniques of Artificial Intelligence (AI), have proved to be suitable in Data Mining for handling this type of series. This paper presents the use of ANNs and Genetic Algorithms (GA) with a time series in the field of Civil Engineering where the predictive structure does not follow the classic paradigms. In this specific case, the AI technique is applied to a phenomenon that models the process where, for a specific area, the fallen rain concentrates and flows on the surface. [ABSTRACT FROM AUTHOR]

Auction markets provide centralized procedures for the exposure of purchase and sale orders to all market participants simultaneously. Online auctions have effectively created a large marketplace for participants to bid and sell products and services over the Internet. eBay pioneered the online auction in 1995. As the number of demand for online auction increases, the process of monitoring multiple auction houses, picking which auction to participate in, and making the right bid become a challenging task for the consumers. Hence, knowing the closing price of a given auction would be an advantage since this information will be useful and can be used to ensure a win in a given auction. However, predicting a closing price for an auction is not easy since it is dependent on many factors. This paper reports on a predictor agent that utilises the Grey System Theory to predict the closing price for a given auction. The performance of this predictor agent is compared with another well known technique which is the Artificial Neural Network. The effectiveness of these models is evaluated in a simulated auction environment. [ABSTRACT FROM AUTHOR]

In this paper, Glucose concentration in urine sample were determined by means of Near-infrared spectroscopy combination with artificial neural networks (ANNs). Distinct ANN models were developed for the multicomponent analysis of biological samples. The developed prediction algorithms were tested on unseen data and exhibited high level of accuracy, comparable to the partial least squares (PLS) regression method. Near-infrared (NIR) spectra in the 10000~4000 cm− 1 region were measured for human urine glucose with various concentrations at 37 °C. The most informative spectral ranges of 4400~4700 cm− 1 was selected by principle component analysis (PCA) for glucose in the mixtures. For glucose, the correlation coefficient (R) of 0.999 and the root mean square error of prediction (RMSEP) of 20.61 mg/dl were obtained. The presented algorithms can be used as a rapid and reliable tool for quantitative analysis of glucose in biological sample, since the concentration can be determined based on NIR measurement. [ABSTRACT FROM AUTHOR]

This paper investigates the use of Artificial Neural Networks (ANN) to mine and predict patterns in software aging phenomenon. We analyze resource usage data collected on a typical long-running software system: a web server. A Multi-Layer Perceptron feed forward Artificial Neural Network was trained on an Apache web server dataset to predict future server swap space and physical free memory resource exhaustion through ANN univariate time series forecasting and ANN nonlinear multivariate time series empirical modeling. The results were benchmarked against those obtained from non-parametric statistical techniques, parametric time series models and other empirical modeling techniques reported in the literature. [ABSTRACT FROM AUTHOR]

In the present paper are described the component stages implemented on the Automatic Speech Recognition System for Romanian Language, ASRS_RL and the speech recognition experiments made with the system. The discussed stages are the feature extraction based on cepstral analysis and linear prediction and the learning strategies implemented on the system for acoustical modelling. The first goal of our research was to develop and experiment a system for continuous speech recognition and understanding in the statistical framework of hidden Markov models (HMM). A second goal was to asses other learning strategies like support vector machines (SVM), artificial neural networks (ANN), and some hybrid structures as alternatives for the HMM. The made experiments in isolated words (digits) and vowels recognition gave good results and proved the competitively of the applied new structures. [ABSTRACT FROM AUTHOR]

Importance of construction of models for predicting software quality attributes is increasing leading to usage of artificial intelligence techniques such as Artificial Neural Network (ANN). The goal of this paper is to empirically compare traditional strategies such as Logistic Regression (LR) and ANN to assess software quality. The study used data collected from public domain NASA data set. We find the effect of software metrics on fault proneness. The fault proneness models were predicted using LR regression and ANN methods. The performance of the two methods was compared by Receiver Operating Characteristic (ROC) analysis. The areas under the ROC curves are 0.78 and 0.745 for the LR and ANN model, respectively. The predicted model shows that software metrics are related to fault proneness. The models predict faulty classes with more than 70 percent accuracy. The study showed that ANN method can also be used in constructing software quality models and more similar studies should further investigate the issue. Based on these results, it is reasonable to claim that such a model could help for planning and executing testing by focusing resources on fault-prone parts of the design and code. [ABSTRACT FROM AUTHOR]

The new type of flywheel battery requires control system with compact structure and low manufacturing cost. To meet this requirement, a new method for the sensorless vector control of flywheel battery is proposed in this paper. The advantage of the proposed control system is that it does not need an extra sensor to obtain the flywheel speed and position information. The determination of flywheel position and thereby speed are made by estimating back electromotive force (EMF) using the artificial neural network (ANN) observers. By doing so, the dimensions and cost of the driver system can be reduced. The ANN observers use the instantaneous values of stator voltages and currents and the estimated error of the stator current as their input to output the back EMF components in the α-β reference frame. A simulation model was established by the use of MATLAB/Simulink software to carry out the numerical experiments. The test results demonstrate that the proposed charging control system for flywheel battery has a good control performance and a good robustness. The speed /position estimation precision is high and the error is acceptable for a wide speed range. [ABSTRACT FROM AUTHOR]

Hand gestures recognition is the natural way of Human Machine interaction and today many researchers in the academia and industry are interested in this direction. It enables human being to interact with machine very easily and conveniently without wearing any extra device. It can be applied from sign language recognition to robot control and from virtual reality to intelligent home systems. In this paper we are discussing work done in the area of hand gesture recognition where focus is on the soft computing based methods like artificial neural network, fuzzy logic, genetic algorithms, etc. We also described hand detection methods in the preprocessed image for detecting the hand image. Most researchers used fingertips for hand detection in appearance based modeling. Finally we are comparing results given by different researchers after their implementation. [ABSTRACT FROM AUTHOR]

To meet the requirement of real time strategy (RTS) games, this paper proposes a two-stage speed up model using the combination of genetic algorithm and ant algorithm (GA-AA) and then artificial neural network (ANN) in a RTS game. The task is to tackle the optimization and adaptive defensive positioning game problems. In the first stage, we use GA to perform the initial optimization, and then AA is incorporated to speed up GA. In the second stage, the results of GA-AA can be used as cases to train an ANN, which obtains the optimal solutions very fast and then completes the whole off-line learning process. These optimal solutions stored in the trained ANN are considered to be still useful to provide good solutions in even random generated environments. Thus the two-stage speed up model not only needs less off-line training time, but also can recommend good on-line solutions very quickly. Experimental results are demonstrated to support our idea. [ABSTRACT FROM AUTHOR]

In classification, when the distribution of the training data among classes is uneven, the learning algorithm is generally dominated by the feature of the majority classes. The features in the minority classes are normally difficult to be fully recognized. In this paper, a method is proposed to enhance the classification accuracy for the minority classes. The proposed method combines Synthetic Minority Over-sampling Technique (SMOTE) and Complementary Neural Network (CMTNN) to handle the problem of classifying imbalanced data. In order to demonstrate that the proposed technique can assist classification of imbalanced data, several classification algorithms have been used. They are Artificial Neural Network (ANN), k-Nearest Neighbor (k-NN) and Support Vector Machine (SVM). The benchmark data sets with various ratios between the minority class and the majority class are obtained from the University of California Irvine (UCI) machine learning repository. The results show that the proposed combination techniques can improve the performance for the class imbalance problem. [ABSTRACT FROM AUTHOR]

Does the way a person read influence the way they understand information or is it the other way around? In regard to reading of English text, just how much we can learn from a person΄s gaze pattern? It is known that while reading, we inadvertently form rational connections between pieces of information we pick up from the text. That reflects in certain disruptions in the norms of reading paradigm and that gives us clues to our interest level in reading activities. In this paper, we validate the above statement and then propose a novel method of detecting the level of engagement in reading based on a person΄s gaze-pattern. We organised some experiments in reading tasks of over thirty participants and the experimental outputs are classified with Artificial Neural Networks with an approximately 80 percent accuracy. The design of this approach is simple and computationally feasible enough to be applied in a real-life system. ˵Your eyes are the windows to your soul″ [ABSTRACT FROM AUTHOR]

In this paper we present a computer aided diagnosis (CAD) system for mass detection and classification in digitized mammograms, which performs mass detection on regions of interest (ROI) followed by the benign-malignant classification on detected masses. In order to detect mass effectively, a sequence of preprocessing steps are proposed to enhance the contrast of the image, remove the noise effects, remove the x-ray label and pectoral muscle and locate the suspicious masses using Haralick texture features generated from the spatial gray level dependence (SGLD) matrix. The main aim of the CAD system is to increase the effectiveness and efficiency of the diagnosis and classification process in an objective manner to reduce the numbers of false-positive of malignancies. Artificial neural network (ANN) is proposed for classifying the marked regions into benign and malignant and 83.87% correct classification for benign and 90.91% for malignant is achieved. [ABSTRACT FROM AUTHOR]

The web international is one of the main sources of knowledge. This knowledge can be texts, images, and photos etc. The experts are always looking for the best way to find the exact information and they want to retrieve it as quickly as possible. But, this information is available on different servers in the form of web pages in different languages. Now, we need to find the most efficient technique which can collect and display the web pages which have similar information within a fraction of a second. The only and easy solution is to do this work in automated way. From this automated method we get many systems of knowledge which are called hypertext, or hypermedia. These methods are used in different domains. Many statistical and mathematical techniques are used to group the web pages which have similar information. In this paper we have proposed an Artificial Neural Network based mathematical model to find the similarity between web pages. [ABSTRACT FROM AUTHOR]

Sequence alignment has been a commonly adopted technique for annotating gene functions. Biologists typically infer the function of a unknown query gene according to the function of the reference subject gene that shows the highest homology (commonly referred to as the ˵top hit″). BLAST search against the NCBI NR database has been the de facto ˵golden companion″ in many applications. However, the NR database is known as noisy and contains significant sequence redundancy, which leads to various complications in the annotation process. This paper proposes an integrative approach that encompasses natural language processing (NLP) for feature representation of functional descriptions and a novel artificial neural network customized based on the Adaptive Resonance Associative Map (ARAM) for clustering of subject genes and for reducing their redundancy. The proposed approach was evaluated in a model legume species Medicago truncatula and was shown highly effective in our experiments. [ABSTRACT FROM AUTHOR]

Existing defect management tools provide little information on how important/urgent for developers to fix defects reported. Manually prioritizing defects is time-consuming and inconsistent among different people. To improve the efficiency of troubleshooting, the paper proposes to employ neural network techniques to predict the priorities of defects, adopt evolutionary training process to solve error problems associated with new features, and reuse data sets from similar software systems to speed up the convergence of training. A framework is built up for the model evaluation, and a series of experiments on five different software products of an international healthcare company to demonstrate the feasibility and effectiveness. [ABSTRACT FROM AUTHOR]

Software reliability is one of the most important software quality attribute and Software reliability estimation is a hard problem to solve accurately. However for management of software quality and standard practice of the organization, accurate reliability estimation is important. Non-homogeneous Poisson Process (NHPP) models and Artificial Neural Network (ANN) models are among the most important software reliability growth models. In this paper we study an approach using past fault-related data with cubic spline Network model to estimate reliability. A numerical example is shown with simulated datasets. The example shows that the proposed model accurately estimate the software reliability. [ABSTRACT FROM AUTHOR]

This paper conducts an investigation on marine diesel engine condition monitoring based on artificial neural network. The engine surface vibration signals are measured and analyzed using time series method. The characteristic parameters of engine vibration signals obtained from time series analysis are used to build a suitable artificial neural network by means of Back Propagation Algorithm in order to detect the engine operating faults and improve the diagnosis accuracy. The diagnoses of variations in valve clearances and engine cylinder loads are discussed. The relationships between vibration characteristics and engine working conditions are presented. Experimental investigations were carried out on a medium speed 6-cylinder marine diesel engine. The cylinder cover vibration signals are measured and analyzed for monitoring the working conditions of valve clearances and engine loads. The results show that it is effective to detect the variations in valve clearance and engine cylinder load from cylinder cover vibration signals. [ABSTRACT FROM AUTHOR]

This paper describes the development of a mechanical fault diagnosis system for a scooter engine platform using continuous wavelet transform and artificial neural network techniques. Most of the conventional techniques for fault diagnosis in a mechanical system are based primarily on analyzing the difference of signal amplitude in the time domain or frequency spectrum. In the present study, a continuous wavelet transform (CWT) algorithm combined with a feature selection method is proposed for analyzing fault signals in a scooter fault diagnosis system. The artificial neural network technique using back-propagation and generalized regression are both used in the proposed system. The effectiveness of the proposed system using two algorithms in CWT technique for scooter fault diagnosis are investigated and compared. The experimental results indicated that the proposed system achieved a fault recognition rate over 95% in the experimental platform of scooter fault diagnosis system. [ABSTRACT FROM AUTHOR]

Artificial neural network (ANN) is a very useful tool in solving learning problems. Boosting the performances of ANN can be mainly concluded from two aspects: optimizing the architecture of ANN and normalizing the raw data for ANN. In this paper, a novel method which improves the effects of ANN by preprocessing the raw data is proposed. It totally leverages the fact that different features should play different roles. The raw data set is firstly preprocessed by principle component analysis (PCA), and then its principle components are weighted by their corresponding eigenvalues. Several aspects of analysis are carried out to analyze its theory and the applicable occasions. Three classification problems are launched by an active learning algorithm to verify the proposed method. From the empirical results, conclusion comes to the fact that the proposed method can significantly improve the performance of ANN. [ABSTRACT FROM AUTHOR]

Controlling chaos in a passive biped robot with an artificial neural network is investigated in this paper. The dynamical model is based on the compass-like biped robot proposed by Garcia et al. (1998) with a point-mass at the hip and infinitesimal point-masses at the feet ignoring the scuffing situation. The governing dynamics and chaotic behavior of the system is explored and the bifurcation diagram is drawn with respect to the ramp slope. Controlling chaos is based on stabilizing the unstable periodic orbits in the chaotic attractor. The UPOs are detected using an iterated algorithm. The artificial neural network is constructed using the information of seven previous steps and the control parameters in each one. The network is trained to find the appropriate control parameter in order to put the next step on the unstable periodic orbit. The control parameter is the toe-off impulse at the heel strike. [ABSTRACT FROM AUTHOR]

This paper gives a brief description of current situation of distributed generation system, and points out that microgrid can run in two kinds of operation modes. The key problems which need to be cautiously considered exist in each operation mode are summarized, and advanced artificial intelligence techniques are adopted to solve those problems as effective tools. The application situation and research status of multi-agent system, artificial neural network, genetic algorithm, fuzzy logic in distributed generation system home and abroad are summarized detailedly. The existing problems and solving thoughts of each artificial intelligence technique are analyzed, and developing directions of artificial intelligence applications in distributed generation system in the future are also prospected. [ABSTRACT FROM AUTHOR]

In this paper, an optimization system is established based on a hybrid neural network and genetic algorithm approach. The application program is compiled in Matlab engineering computing language, which is used in calculating the parameter value predicted by neural network and the result of genetic algorithm optimization .The comparison and error analysis has been carried out between the results predicted by network and CAE simulated results, which shows that the BP network is stable and reliable. The optimized outcome verified by CAE simulation and tested by experiment has been proved to be correct. It has been bean indicated that the injection parameter optimization method based on the hybrid neural network and genetic algorithm approach is feasible. [ABSTRACT FROM AUTHOR]

This paper presents the Static Synchronous Compensator΄s (StatCom) voltage regulation by a B-Spline neural network. The fact that the electric grid is a non-stationary system, with varying parameters and configurations, adaptive control schemes may be advisable. Thereby the control technique must guarantee its performance on the actual operating environment where the StatCom is embedded. An artificial neural network (ANN) is trained to foresee the device΄s behavior and to tune the corresponding controllers. Proportional-Integral (PI) and B-Spline controllers are assembled for the StatCom΄s control, where the tuning of parameters is based on the neural network model. Results of the lab prototype are exhibited under different conditions. [ABSTRACT FROM AUTHOR]

Breast cancer is the second leading cause of cancer deaths in women today and the most common cancer among women. At present there is no known method to prevent breast cancer but early detection increase the chance of cure. Screening mammograms is considered as the best tool for doctors to detect breast cancer at an early stage. In this paper we present the application of Classification based on multiple association rule (CMAR) in neural network (NN) to classify breast cancer mammographic data. CMAR is used in the initial step in creating structure of neural network. It is tested on the real datasets: Mammography Mass Data from UCI repository. [ABSTRACT FROM AUTHOR]

Development of real time strategy game AI is a challenging and difficult task because of the real-time constraint and the large search space in finding the best strategy. In this paper, we propose a machine learning approach based on genetic algorithm and artificial neural network to develop a neural-evolutionary model for case-based planning in real time strategy (RTS) games. This model provides efficient, fair and natural game AI to tackle the RTS game problems. Experimental results are provided to support our idea. This model could be integrated with warbots in battlefields, either real or synthetic ones, in the future for mimic human like behaviors. [ABSTRACT FROM AUTHOR]

This paper illustrates a conceptual framework for evaluating the impact of contemporary information systems mainly on business events. In the process of developing new data mining techniques and software to aid in better business solutions, we propose an 'efficient information system' that hybrids a tool of data mining incorporating rule induction on crisp database which is then transformed into a fuzzy database on which all samples Apriori tree search algorithm generates all possible prominent cases which are finally calibrated using a single layer feed forward rule based artificial neural network to make suggestive predictions and hence decisions. This model is hybrid to serve the aim of designing system architecture highly intelligent, both in system side aspects as well in the applicatory aspects. The system design is worked on an example case to deal with the customer risk management in the area of investment in the stock market. [ABSTRACT FROM AUTHOR]

We synthetically applied computer vision, genetic algorithm and artificial neural network technology to automatically identify the tomatoes that had physiological diseases. Firstly, the tomatoes΄ images were captured through a computer vision system. Then to identify cavernous tomatoes, we analyzed the roundness and detected deformed tomatoes by applying the variation of fruit΄s diameter. Secondly, we used a Genetic Algorithm (GA) trained artificial neural network. Experiments show that the above methods can accurately identify tomatoes΄ shapes and meet requests of classification; the accuracy rate for the identification for tomatoes with physiological diseases was up to 100%. [ABSTRACT FROM AUTHOR]